I hope you find this essay as inspiring as I do.

Reclaiming a Vision of Humane Schooling

by Alicia Stoller

In January of 2021, as we emerged from the darkest phases of the pandemic, the rabbi and climate activist Shoshana Meira Friedman wrote about the challenges of raising children with love and hope amid so much loss. When her toddler son, who had fallen in love with nature while the world was shuttered, learned that a grove of hemlock trees he’d named “Frog and Toad’s Corner” was dying, Friedman wrote:

“I could scarcely breathe, startled by the sudden clarity that I am teaching my child to love a dying and transforming world, that he will learn to love and lose in the same breath, and that I will learn along with him.”

As our devices buzz against our skin each day, warning of one potential existential threat after another, raising and teaching children has become laced with growing anxiety.

It is hard not to let these worries consume us.

The values that inform our teaching, indeed the purpose of education itself, also become more contentious during times of rupture. The function of education is typically defined by the dispositions and skills that adults believe children will need to thrive in an imagined future. On its face, this makes sense. If the most basic purpose of school is to facilitate growth and to prepare children for adult life, it is logical to fix our gaze on what’s to come.

But when fear overwhelms us, preoccupation with the future can become a trap. And in trying to protect our children, we may find we’ve forfeited their childhood.

Pervasive fear of the future also creates openings for predatory versions of school, built to profit off our anxiety. When we view education solely as armor that will protect our children from hardship or scarcity, we risk sacrificing their full humanity today, and the possibility of a less dire tomorrow.

But, like Friedman’s son hugging his beloved hemlock trees, children deserve the fullness of the present. So what might it look like to reject fear, and instead reclaim a humane education that insists upon the dignity of our children today, and the possibility of a more hopeful tomorrow?

Here is what I think it will require.

Step 1: Unmasking the Wolves

Having spent over two decades in both independent (private) and public school classrooms, as a teacher, a school leader, a student-teaching advisor, an educational consultant, and a parent, I have found that most schools espouse similar values. They are the words that flash across school websites and emblazon bulletin boards: curiosity, creativity, compassion, resilience, wonder, love of learning, and student agency, to name just a few.

In many schools, these values are deeply lived, as teachers model them and bring them to life for students each day. Sometimes, though, articulated values only serve to provide a virtue-signaling haze meant to obscure less virtuous goals — usually some external profit or power.

The history of education is littered with examples of children being used as cover for such schemes. This would not be possible if the engineers of these efforts flatly admitted that they don’t really care about children. So, like the wolf in the fairy tale, they disguise their intentions in words we trust. We are seeing examples of this now in ed-tech companies that claim to be guided by notions of literacy, while actually cultivating future consumers, and in schools that claim to nurture innovators, while actually fostering acceptance of predation.

Alpha School, for example, asserts that its educational model is “designed for joy.” Its website declares that instilling “a lifelong passion for learning,” and ensuring that “every student feels valued and motivated” are core values. The rift between these stated values and Alpha School’s actual teaching practices is jarring. Monitoring a student’s heart rate while they are publicly criticized by their parents, or using software to track and correct a child’s emotions, does not instill joy or make kids feel valued. I suspect that the parents who have been drawn into these schools are compelled, at some level, by their own fears of an uncertain future; the only explanation for turning school into “boot camp” is a belief that the path ahead will be combative.

Powerful people want us to believe that our kids’ survival depends on accepting callousness and inhumanity. But I am not willing to forgo our children’s dignity. Nor should any educator or parent.

Step 2: Reclaiming Humane Educational Values

Education is fundamentally an optimistic endeavor. As we work to contain the wolves, we need to articulate a more hopeful version of school. Here are a few anchor points for a humane vision of schooling that lives up to its stated values:

• A joyous spirit of exploration
  Yes, let’s start with sincere, authentic joy. Joyous classrooms begin with teachers who delight in the content of their curriculum, and in the minds of their students. Let’s nurture teachers who relish knowledge and revel in the questions, curiosities, and insights of children.

• A wild embrace of wonder and imagination
  The deepest learning begins in wonder, not in the cold pursuit of achievement. Let’s cultivate classrooms that spark wonder at the human spirit and at the mysteries of the universe in equal measure: views of the stars, muddy adventures, and poetry that opens our students’ hearts and minds.

• A deep and abiding empathy
  There is a reciprocal relationship between care and learning. Let’s look for invitations to empathy in every element of the school day — in the dilemmas of history, the journeys of fictional characters, the diverse vantage points of other beings in the natural world, and in students’ own connections with one another. Let’s fill children’s days with a kaleidoscope of perspectives. Empathy grows from frequent, meaningful encounters with diversity.

• A willingness to work diligently and with fortitude
  Learning often requires persistence and stamina. But a tolerance for productive struggle doesn’t have to operate in opposition to joy, wonder, empathy, or a child’s own dignity. We can do this without stripping students of their self-worth. Let’s encourage students to grapple with challenging ideas and skills, motivated by their own drive and cheered on by our vociferous belief in their capacity.

• A tolerance for uncertainty & patience in pursuit of wisdom
  Learning involves engaging with complexity, and uncertainty is often a sign that we are on the edge of wisdom. Let’s give students opportunities to sit with the discomfort of the unknown. And let’s offer them warm company in such times, so they know that adults are not immune to confusion or doubt themselves.

• Love for the world as it is, and as it could be
  It is possible to learn about the world without loving it. But learning without love, even when love breaks our hearts, diminishes children’s humanity and abandons the full range of possibilities for their future. Let’s cultivate opportunities for children to hug hemlock trees, cradle oysters, gaze at Juliet on the balcony, and “praise the mutilated world.”

The way in which we value children as they stand before us instructs them in who they are permitted to become, and in the values they should uphold themselves. When our fear of the future obscures the beauty of a child’s developing, grappling, sensitive, inquisitive mind, it overwhelms their essential humanity — and our own.

We can do better.

Children are paying constant, close attention to the world around them. The values we espouse for their future are only credible when our actions, as the adults in their lives, bear those values out in the present. When we delight in the questions, curiosities, imagination, and even in the confusion of children, we stand guard against those who claim to speak for their future. We contain the wolves. And we give children back the right to imagine their own future.

As the poet Mary Oliver implored on behalf of children:

“Give them the fields and the woods and the possibility of the world salvaged from the lords of profit.”

The future will be our children’s to navigate. But the values they carry into that future will only be as sturdy as our willingness to embrace their growing hearts and minds in the present, and to hold fast to a vision of a decent and empathetic world — one they will create simply by inhabiting it.

Love for the world as it is, and as it could be. Our schools, all of our schools, must reclaim and redeem this humane possibility.

My thanks again to Alicia Stoller for writing and sharing this elegant testimony to the beauty and wonder of education. We plan to collaborate further to articulate the possibilities of humane schooling. And I invite those who might be interested in joining in that project to let us know in the comments. We will contain the wolves, and we will reclaim the future that our children deserve. Education proclaims a human will.

Subscribe now

I am reporting three important pieces of new information related to Alpha School’s Survival Challenge that have now been confirmed by multiple sources.

I have run out of words to describe the depth of my concern for child safety at Alpha School. I will keep this brief (at least by my standards), describing these new facts before providing brief commentary.

1. Students were deprived of their life-saving medical supplies for the duration of the Survival Challenge

Alpha School put its students’ lives in danger during the Survival Challenge by confiscating their medical supplies.

According to several sources, on the first day of the Survival Challenge, all students were required to turn over their medical supplies to adult guides. These sources inform me that these medical supplies were not returned to students until Sunday at the conclusion of the event. (Several students didn’t get their supplies back until Monday, according to one source.)

Why Alpha School did this is unknown. What is known that is some of the medical supplies that students were deprived of over the weekend included epi-pens and inhalers, according to multiple sources.

This was (and is) extraordinarily dangerous. An epi-pen injects epinephrine to quickly address severe food allergies known as anaphylaxis. According to the Food Allergy Research Education organization (FARE), for people with such allergies, it is “crucial to have epinephrine with you at all times” (my emphasis). Indeed, FARE recommends carrying two epi-pens, aka epinephrine auto-injectors, at all times as a safety precaution.

The consequences of failing to immediately address a serious allergic reaction are severe; according to the American Academy of Allergy Asthma and Immunology, these reactions are such that they “typically come on quickly and may cause death” (my emphasis). The AAAAI further notes, “this medical emergency requires immediate treatment…if you (or anyone you are with) begin to have an allergic reaction, use your epinephrine and get to the closest emergency room. The sooner the reaction is treated, the less severe it is likely to become.”

For three days and two nights, Alpha School students were deprived of access to their epi-pens. While the severity of students’ anaphylaxis conditions are unknown, there is no justification — none — for any policy or practice that results in students lacking access to this potentially life-saving medical device.

Likewise, there is no possible justification for depriving students of access to their inhalers, either. The National Heart, Blood, and Lung Institute states that asthma attacks “can be very scary” and that someone suffering such an attack “should use your reliever inhaler as soon as you start to have symptoms.”

Alpha School put allergic and asthmatic students in danger for the duration of the Survival Challenge. I do not know what other forms of medicine might have been confiscated, but my investigation into this remains ongoing.

2. Student attendance was not taken during the Survival Challenge

I reported earlier this week that 73 high school students attended the Survival Challenge, a figure supplied to me by multiple sources.

What I have since learned through multiple sources is that Alpha School did not take the attendance of students once the Survival Challenge commenced — students were accounted for at the beginning of the event and when it concluded, but not in between. Given the 250-acre size of the River Rock Ranch, its minimal gating and fencing, and the separate campsites of adult guides and students, it is fair to say that students’ presence on the ranch went unaccounted for long durations within the Survival Challenge time period.

This is dangerous and unsafe. The Survival Challenge started on a Friday and ended on a Sunday (albeit earlier than planned, according to multiple sources). A student could have wandered through the wilderness and seriously injured themself while away from others, yet could have potentially remained unaccounted for up to 48 hours. This is why so-called normal schools take attendance with regularity: at all times, public schools must be able to account for the whereabouts of the children in their care in order to keep them safe. According to Texas Education Agency memo dated February 2, 2026 on Attendance, Complaints and Educator Responsibilities1:

• “Schools have a responsibility and duty to keep students safe.”

• School systems have an “obligation to ensure students attend school and remain safe.”

• State law recognizes parental involvement as including “parents knowing the location of their child during the school day.”

• And schools “should make every effort to notify parents when students may leave school property.”

So that’s state law that applies to normal schools in Texas. But then, Alpha School does not consider itself a normal school.

3. Circumstances surrounding the adult guide fired in relation to the Survival Challenge

Additional information related to the Alpha School guide that was fired has come to light.

Through multiple sources, I’ve learned that not long after arriving at the River Rock Ranch, one adult guide stated they were suffering heat stroke. Sources indicate that other staff disagreed with this self-assessment. Sources further indicate this guide then left the ranch of their own volition to visit a hospital in Victoria, Texas, and obtained a note from a doctor that stated the guide should no longer stay on the ranch, and should instead return home. The guide did so.

The guide was fired by Alpha School shortly thereafter. I am withholding the name of this guide to preserve confidentiality.

My commentary

Alpha School put the lives of its students in danger, seemingly for no reason other than to cosplay at some twisted version of the reality television show Survivor. The name of the event itself, Survival Challenge, seemed overly dramatic upon my first reading, but based on what I’m reporting today and have reported previously, it appears there was a non-trivial danger that students could have suffered a major medical crisis, or even died, while participating in this activity at the River Rock Ranch (owned by Little B. Diddies LLC).

We know that one student fell and was taken to the ER, itself a major concern. But it could have been so much worse.

Finally, I note that Alpha School high school students take many class trips together — homecoming, for example, will be in Miami this year, according to multiple sources. Student field trips to foreign countries also take place, such as Poland. I simply do not know how any parent could feel safe about their child’s participation in these trips, given the inexplicable confiscation of students’ medical supplies, and the failure to take regular attendance to monitor students’ whereabouts.

I ask again: What will it take?

The scandals at Alpha School continue to unfold at dizzying pace.

In what follows, I report on recent events taking place within Alpha School’s boot camp, the first eight-week phase of its high school. What I’ve learned moves beyond being weird or troubling — students were harmed, physically harmed, as the weekend unfolded. There is an open and active police investigation. And student privacy is being violated at scale.

I have organized this report as follows:

1. Initial descriptions of the Survival Challenge

2. Incident during the Survival Challenge that resulted in student hospitalization

3. Open and active investigation by Victoria County police

4. Other troubling incidents during the Survival Challenge

5. The mysterious River Rock Ranch owned by Little B. Diddies LLC

6. Alpha School’s continued silence

7. My commentary

8. Field notes from my trip to Victoria, Texas

What you read may unsettle you, and it should.

Initial descriptions of the Survival Challenge

As I reported in my initial special report, Alpha School’s high school boot camp involves a number of unusual activities. One activity that multiple students described on social media at the beginning of boot camp involved a multiple-day outdoor camping trip, stylized as a Survivor-style event, based loosely on the long-running television “reality” show.

One student called this Alpha School’s “Survival Challenge,” a term I will use throughout this report.1 This student stated that Alpha School would be “dropping [students] in the woods with a single bag of rice and dry beans to see if we can survive.” Further, “if your fire goes out or tarp collapses in the rain, the guides will not fix it for you.” In sum, “instead of sitting in a classroom and doing nothing…Alpha drops you in the wilderness to see how you handle real pressure.” Other student tweets echoed this description of what was planned.

Alpha School itself describes this activity under as cultivating the “life skill” of “leadership,” wherein students “help build a culture where it’s cool to work hard.” The boot-camp microsite also confirms when the Survival Challenge was scheduled to occur:

The Survival Challenge boot-camp event just took place as scheduled, from September 18 to September 20, 2026, a little over a week ago as of the publication date of this report. According to multiple sources, the entire Alpha School high school, approximately 73 students, first gathered in Austin before being transported by bus to the River Rock Ranch in Victoria, Texas for the weekend.

The good news is that every student appears to have actually survived the Survival Challenge. The less good news is that one student fell and was concussed, and had to be provided advanced life support services on-site before being evacuated by ambulance to the local ER. Further, the Victoria County Sherriff’s Office has confirmed there is an open and active police investigation underway. One Alpha School guide was fired as all this unfolded. And there are several other incidents that raise questions about Alpha School’s oversight and ability to ensure student safety and privacy.

I will now describe what I’ve learned about what happened during the Survival Challenge, starting with the general atmosphere and then turning to particular incidents, all confirmed through multiple sources.

Incident during the Survival Challenge that resulted in student hospitalization

The student who tweeted about being dropped in the wilderness with de minimis adult supervision was prescient.

In general, during the Survival Challenge, sources report that students were largely left unsupervised for three days. Although adult guides were present at River Rock Ranch, they set up camp separately from the students, approximately 20-30 minutes away by foot. Students were told that they could go wherever they wanted within the expansive, 250-acre campground. Although there were numerous port-a-potties, sources told me there was only one shower facility.

Multiple sources confirm that at least one major traumatic incident took place over the weekend, one that involved a student falling and becoming concussed. Emergency Medical Services and Fire Department teams were dispatched, and the student was provided advanced life support services on site. The student was then evacuated from the campground by ambulance, and taken to the emergency room at the local hospital.

I know this in part because on Friday, September 26, 2026, I filed a public record request with the Victoria Fire Marshal’s office, and later that day I was provided an incident report for the night of September 19, 2026. The first page of this report appears below, with my circles in red:

Despite repeated attempts, I was unable to speak with any EMS technician based in Victoria to learn more about what Advanced Life Support services they might typically provide to a concussed juvenile. However, this Texas Emergency Medical Services Initial Provider License Checklist describes ALS services generally, with my highlighting:

As one official from the Victoria Fire Marshal’s office said to me, ALS services are “more than just a band-aid.”

Open and active investigation by Victoria County police

The local police are investigating what happened over the Survival Challenge weekend.

On Friday, September 26, 2026, I filed a public record request with the Victoria County Sherriff’s Department seeking “any public record that memorializes any dispatch by the Police Department or law enforcement provider to River Rock Camp over the time period in question (9/18/26 to 9/20/26).” The afternoon of Monday, September 28, 2026, I received the following notice, with my highlighting:

I spoke with Erika Ruiz, the records clerk, that same afternoon, and she confirmed it is standing policy in Victoria County to not release any records related to an active police investigation. She added that, “since they are juveniles, that also comes into play, since you’re not their parent.” She confirmed that the statement regarding “relying on of the Office of the Attorney General’s previous determination” refers to this Open Records Memorandum Rulings from the OAG, which in turn states that “juvenile law enforcement records are confidential” and “may not be disclosed to the public,” but may be produced pursuant to a request from a child’s parent or guardian.

So all we know right now is that there is an open and active investigation by Victoria County police related to the Survival Challenge on the River Rock Ranch. Parents with students in Alpha School’s high school may be able to obtain the existing police report, and I encourage concerned parents to do so here. I will update this report if additional information related to this police investigation comes to light, while of course always protecting student privacy.

My enduring thanks to those who have pledged to support paid subscriptions here — but I’m asking for a different sort of help via donations to my newly created Cognitive Resonance Ko-fi page, which you can also navigate to by clicking the logo below.

Other troubling incidents during the Survival Challenge

On its own and without further detail, someone falling and being concussed during a camping trip could be understood as just an accident — these, of course, do happen. But leaving approximately 73 teenage students largely unsupervised by adults on a 250-acre campground for multiple days obviously increases the likelihood of such an accident occurring. Sources indicate that at least one guide was fired as the events of the Survival Challenge weekend unfolded, although the precise circumstances and reasons for this are presently unknown (I am actively investigating).

The fall that resulted in a concussed student should not be viewed in isolation, but rather as just one of many concerning things that happened during the Survival Challenge. Sources indicate that many students suffered from sunburn and/or heat exhaustion. Texas is experiencing a hot fall, and Victoria hit highs of 100 degrees over the weekend in question. One source reported to me that, as a planned activity, students were made to walk ¾ of a mile without any water, which was only provided upon completion of this stage of their hike.

In addition to student safety, there are also troubling issues of student privacy violations. As noted, there was only a single shower facility that the 73 students had access to for three days. Multiple sources informed me that this facility was not fully private, and that male students repeatedly observed female students showering through gaps in the shower’s walls.

Finally, although my sources informed me that adults were largely absent during the Survival Challenge, this does not mean student went unobserved. Multiple sources indicated that Go Pro cameras were placed around where students camped, and positioned to face the area where students slept at night. I’m also told drones were omnipresent.

For broader context, I reached out to my longtime friend and collaborator Dylan Kane for perspective on Alpha School’s Survival Challenge. Although Dylan has recently received deserved attention for his reduced use of screens in his classroom, he also taught for six years at the High Mountain Institute in Colorado, an outdoor expeditionary school, and he’s spent over 40 weeks leading wilderness trips with students ranging from 15 to 18 years old. He told me this:

What Alpha School is doing has some things in common with practices in the outdoor education world, where students camp or travel with minimal adult oversight. Two things to note: first, when we did things like this, they occurred at the end of the program. This was after students had chances to practice the relevant skills with plenty of support, with the goal of creating a meaningful culminating learning experience.

And second, we still had persistent safety issues. My organization, and several other similar outdoor education organizations, scaled back practices like these because we struggled to manage risk and avoid incidents, and because we felt we could get similar learning outcomes with more adult oversight.

Finally, I am reporting here only the incidents that I can presently confirm through multiple sources (apart from the hiking incident that one source agreed to share with me on background). I have reason to believe other concerning things happened during the Survival Challenge weekend, and I will update this report if and when I can confirm them through multiple sources.

I turn now to the River Rock Ranch and what I have — and have not — been able to determine about it serving as venue grounds for the Survival Challenge.

The mysterious River Rock Ranch owned by Little B. Diddies LLC

The website for the River Rock Ranch is unusual. It includes promotional statements about being a blank canvas with no restrictions, a montage of pictures of the property, a short promotional video, and a map showing its general location.

What it does not include is a physical address or phone number — the only public-facing means of communicating is through a web form. Moreover, although there are social media accounts for the ranch, most do not appear to have been active since 2023 (save the LinkedIn account, which has a single fireworks video posted to it from July 2026, stating, “happy birthday America.”)

In the course of my investigation, I spoke several times with a representative from the Victoria County Fire Marshal’s office. During one of our conversations, an official informed me that, according to internal records, the River Rock Ranch is owned by an LLC located in Austin, Texas. He further informed me that the name of the ranch’s legal owner is “Little B. Diddies.”

I have confirmed this. According to Texas public records and online sources, Little B. Diddies is owned and managed by Beau and Shelby Curtis, and the LLC is registered to an address in Austin, Texas. The public social media profiles for the Curtis’s in turn confirm their co-ownership of River Rock Ranch:

I reached out to Beau Curtis and Shelby Curtis via social media to inform them of my reporting, and asked them to explain the Little B. Diddies relationship with Alpha School, but they did not respond.

Texas is well-known for its devotion to limiting government regulations, and it is unclear what regulatory rules or local ordinances may apply to the River Rock Ranch. I spoke with several officials at various departments at both the City of Victoria and the County of Victoria, including the respective fire marshal, planning and development, and engineering departments, to determine what official records might exist related to the Ranch.2 But apart from identifying Little B. Diddies LLC as its legal owner, these officials were unable to find any public records related to the property at all.

Adding to the complexity, the address where EMT and fire department dispatched to, 670 Smith Road, falls outside the City of Victoria’s jurisdiction, but two officials told me it might fall within a special area called the ETJ, for “Extra Territorial Jurisdiction,” that exists within a 3.5 mile radius around the City of Victoria. The same officials told me there were no records associated with River Rock Ranch within that jurisdiction, either.3

One final note on the River Rock Ranch: On Saturday, September 27, 2026, I drove to 670 Smith Road in Victoria, Texas, approximately two hours from Austin. A video and two pictures of what I observed are posted below, and I provide further personal commentary and observations in the field notes section of this report.

Alpha School’s continued silence

On the evening of Saturday, September 26, 2026, I emailed Anna Davlantes, Alpha School’s Chief Communications Officer, to again offer Alpha School the opportunity to comment on my forthcoming story about the Survival Challenge. I attached a copy of the fall incident report and requested comment by end of day Monday, September 27.

I received no response.

Concurrently, I emailed the same offer to Chris Locke, Alpha School’s Head of High School, who’s held that role since June 2021, at least according to his public LinkedIn profile:

In my email, I stated the following: “Given the nature and seriousness of what I plan to report, I want to extend the opportunity to comment directly to you as the person who I understand oversees Alpha School’s high school, and therefore bears direct responsibly for the ‘Survivor Challenge’ and any harm that resulted to students.”

Locke did not reply. But for context, I want to share two additional pieces of information stemming from my general investigation of Alpha School.

First, Chris Locke’s name has come up in several conversations with sources over the past weeks. Without exception, even sources that are highly critical of Alpha School have described him favorably, and as someone committed to educating the students in his care. As I stated to Locke in my email, I have long believed that good people can work for bad organizations.

Second, multiple sources have told me that Joe Liemandt pays particular attention to Alpha School’s high school, and that many activities within it are created as the result of his direct instruction to staff. As one source told me, these activities often reflect “whatever the last thought is that entered Joe’s brain.” It is notable that in addition to the Survival Challenge, high school students in boot camp are also tasked with learning “critical thinking” by watching a full season of Survivor: Cagayan.

Whether Joe Liemandt is a fan of Survivor is not something I plan to investigate at this time, but I suggest anyone who interviews him in the future put the question directly to him.

My commentary

Alpha School is in complete crisis. Children are being harmed. The police are investigating. And the false narrative that its leadership has long presented to the world is destroyed forever.

I now want to speak directly to several affected communities.

To students enrolled at Alpha School: You have the right to be safe at Alpha School. You also have the right to personal privacy at Alpha School. If you are asked to do anything by anyone at Alpha School that you feel violates these rights, you should object, document your objection, and refuse to do what you are being asked. You should tell your parents or guardians immediately. You do not have to do anything that puts your physical well-being at risk. Nor do you have to sacrifice your privacy to Alpha School’s PR machine. Do not let them control you to the detriment of your safety or privacy.

To parents with students enrolled at Alpha School: In the wake of my reporting, many people have asked why you continue to keep your children enrolled at Alpha School, and I have no good answer. In my previous reporting, I stated my belief that children are at risk. With alarming speed, the events of the Survival Challenge have now borne that out. I am speaking to you as much as a fellow parent as I am citizen journalist: your children are not safe attending Alpha School.

To adults employed by Alpha School: Based on multiple sources and conversations I’ve had throughout my investigation, I understand you have received little or no training related to ensuring the safety of the children in your care. Nonetheless, as a matter of common decency and common sense, you should exercise oversight to ensure of your students, and to protect their privacy.

I know many of you are afraid of Joe Liemandt. If that’s the case, resign. Your moral obligations are to a power higher than any he will ever possess.

To Joe Liemandt and MacKenzie Price: You are the co-founders of Alpha School and bear ultimate responsibility for its activities. The same guidance I just provided to the adults employed by Alpha School apply with far more force you both. Students, which is to say children, are not guinea pigs or lab rats for you to experiment upon, nor are they proto-influencers to manipulate, shredding student privacy in the process, so as to generate footage to promote the hideous organization you purport to lead.

I am told by many that you view Alpha School as your legacy project. Whatever remains of your already tattered legacy will be irreparably destroyed if you continue to harm children due to your neglect of their safety, and their privacy.

I am so ashamed of you.

To the grotesque and terminally online pro-Alpha School nitwits I call the Alpha-Betas: For several weeks, you have defamed my character in social media. I have paid you little direct attention, because I’ve been busy reporting on actual things happening in the real world that violate the rights of students at Alpha School, things that are actually creepy. The baseless claims you’ve circulated among your sociopathic crew do not intimidate me. You are nothing to me.

I have little doubt you will ignore the harms to actual children I share in this report, but perhaps this will get your attention: Alpha School’s bumbling communications team has been relaying your missives to news outlets both big and small. This means that your disgusting missives are now intertwined with the Alpha School empire, and as such, you and Alpha School may jointly bear responsibility for the baseless smears you’ve so glibly shared on social media. I already have general counsel monitoring your activity, and I am in the process of obtaining litigation counsel to affirmatively pursue my claims. If you do not cease your defamatory posts, you will soon need to lawyer up.

Continue to fuck around, and you will find out.

To the education community broadly: What is it going to take? What is it going to take? I thought by now I’d see more people jumping off the rapidly derailing Alpha School hype train, but instead, the reaction remains muted among those who should know better. What should be happening is neatly modeled by educator Wess Trabelsi, who recently published a post titled, with admirable candor, “Boy, was I wrong about Alpha School.” Yes! Hell, everything about Alpha School is wrong, and until I trained my attention upon it, I had no idea myself. There is no shame in admitting to being deceived by the carefully crafted but total bullshit story they’ve been promoting for years. Matt Seybold also does outstanding work here pulling up to the broader aims and implications of what Alpha School’s owners ultimately aim to do.

In glaring contrast, what should not be happening are essays such as this one by Michael Horn and Thomas Arnett of the Christensen Institute that gloss over the reality of Alpha School. Their tone-deaf post contains a weak-ass disclaimer regarding the “recent wave of reporting on Alpha School,” before the authors quickly proceed to vomit up a word salad about school choice or whatever. Since these two have been wrong on virtually everything they’ve predicted about the “future of education” for decades, I will spend zero time interrogating their inane arguments, but given that Horn platformed MacKenzie Price and Alpha School on his podcast just last year, I find it unfathomable he chose to publish this essay now. It’s akin to having interviewed Elizabeth Holmes, then watching her get indicted, yet concluding, “well my, that certainly is something, but let me now explain why we still need to disrupt home blood testing.” It’s as clueless as it is embarrassing. (I offered Horn the opportunity to comment on this, but he declined.)

Now let’s try to find something uplifting — and delicious — stemming from all this, shall we?

Field Notes from my trip to Victoria, Texas

Here is something to know about me: the Riley family roots run deep in Texas. Outside of Llano, not far from the local public high school, there is a hill that locals adorably fondly refer to as “Riley Mountain,” so named for my great-great-great-great grandfather James C. Riley, who settled there with his family in 1841. Here’s an old photo of them, courtesy of the friendly folks at the Llano County Historical Museum:

In the generations that followed, my great-great grandfather David W. Riley became a man of some importance in Texas, serving as sheriff of Llano County and also one of the early Texas Rangers, not the baseball team but rather, the paramilitary outfit revered by Texans generally despite it being — or perhaps because it was — often a tool of white supremacy. Relatedly, David W. Riley also joined the Confederate Army, and named one of his five children Robert E. Lee Riley, also known as my great-grandfather (never met him). And it turns out that David W. Riley retired in Austin, and he’s buried about 15 minutes from where I sit typing this now. I’ve visited his grave multiple times.

Anyway, there are many things that drive me bananas about my home state — the right-wing politics being one, the lack of public land another — but I kind of adore enjoy road tripping through its vast if largely privatized expanses.

And boy do I love our barbecue.

So, while investigating the River Rock Ranch in Victoria, Texas — reminder, that’s the one owned by Little B. Diddies — I asked locals to recommend the best barbecue in town. (Pro tip: You can do this just about anywhere in Texas, and you will rarely be disappointed.) There was one and only one that was universally recommended: Mumphord’s Place.

What I am about to say I do not say lightly. After living in Texas for more than a decade and exploring it from the northern panhandle to the southern Rio Grande Valley, from the western border town of El Paso to the eastern woodsy surrounds of Lufkin and Tyler, after conducting my own personal quest to find the best barbecue in a state that reveres smoked meats with religious fervor, I say this without reservation: Mumphord’s Place is the best barbecue joint I’ve found in all of Texas.

I don’t have time to write a proper review, and this report is long enough as it is, so I’ll just quickly share that:

• The brisket consistency was perfect, with a deep crimson smoke ring;

• The ribs were also “competition-grade” perfect, meaning the meat clung to the bone but immediately separated when bitten into (and please trust that I devoured them quickly);

• The sausage was as flavorful as any I’ve ever tasted;

• The beans and potato salad were likewise outstanding (the quality and options of barbecue side dishes are their own separate issue in Texas, believe you me); and

• The banana-foster pudding, oh my god.

The food alone makes the journey to Mumphord’s Place worth your time, but there’s something just as important: the atmosphere. It’s a family affair there, I believe Ricky and Keith Mumphord are the co-owners and pitmasters, and everyone working inside appeared to be either family or close friends, people who grew up together in the same community and deeply know their personal histories, people who clearly love each other, and also love talking endless shit, a joyous form of razzing emanating from back in the kitchen but directed at times toward the seemingly endless line of local customers, too.

That includes me. After Ricky Mumphord kindly let me take a few pictures of him carving up some brisket, the tireless and wonderful server LaJuana Strait rightly noted that women aren’t often seen in BBQ pics because they are more likely to be working the tables rather than slicing the meat. Duly shamed, I vowed to immediately remedy that fact, and she’s pictured below. Hers is a beautiful soul. That’s true of everyone working there, you can feel it the minute you step inside.

I say again: Mumphord’s Place is the best barbecue in Texas. I’ll be coming back soon, y’all.

What, you thought we were finished? Here are three more pictures and brief commentary arising from my road trip to Victoria, Texas.

I reference in my video clip above the sort of ranches I typically see used for group events in Texas. I have no idea if this one is public or private, but this is what I mean about them usually having a nice gated entryway. No animal skull, though – might be owned by a Californian.

Mandatory, and I believe our European World Cup visitors are still recovering from the overwhelming experience that is visiting a Buc-ee’s. While obtaining provisions in Bastrop, I saw one patron staring bewilderedly at its wall of jerky and seemingly endless varieties of seasonings. The move here is the Bohemian Garlic, trust me on this. But make sure to have some floss handy.

I’ll wrap things up now, finally, but one habit that can irk my road-tripping companions is my insistence on stopping to read the many historical markers that line the Texas state highways. Learning our history is joyous for me, even when it reminds us of the many manifestations of white supremacy in this state, as this marker does. We cannot change the past, but we for damn sure can learn from it, and take action to remedy the inequities arising from whence we came.

Take the time to stop and read these geographic footnotes to our history — they are educational and important.

Subscribe now

Longtime readers of this newsletter know I’m a bit obssessed with cultural cognition, and how and why we humans are so unlike the rest of the animal world, including our primate relatives. Through the mysterious workings of social media, several months ago I befriended Dr. Claudio Tennie, who has spent his professional career investigating this very question. I’m not not jealous.

Specifically, Tennie is a professor at the University of Tübingen (Germany) where he holds a group leader position in “Tools and Culture among Early Hominins.” He is trained in biology (especially behavioral biology and primatology), psychology (especially comparative and developmental psychology), archaeology (especially cognitive archaeology and early stone tools) and philosophy (especially philosophy of science).

As you’ll soon read, Tennie’s research explores the question of how humans diverged from our primate relatives (time “A”) to become what we are today (time “B”). Much of our discussion focuses on the relative capacities of humans and great apes to copy from each other, what other cognitive scientists might call imitation, viewed largely through the lens of tool making. We also discuss the role of language in fostering culture and at the end, as always, we touch on AI.

Benjamin Riley: Your social media profile poses a deep question – how did we humans get from an ape-like baseline to human cultural learning and cultural cognition? What is the scientific answer to that question?

Claudio Tennie: Excellent question.

So, this is really the result of me iterating, trial and erroring a lot of possibilities. The big questions are: How do we get from A to B? Why did great apes and humans diverge?

To answer, we have to go back in time. We have to understand where apes come from generally, and then understand the ‘history’ from that point onward.

But the issue is that we don’t have a time machine! We can only very indirectly infer what was going on back then.

We have paleo-anthropology, essentially looking at bones. We have lithic archeology, looking at stone tools. But that is all extremely indirect. And in many ways, they provide unsatisfying data.

I’m a primatologist, and I mainly work with apes in captivity, but I’ve done some gigs out in the wild with several species, not always apes, sometimes birds, and it’s such a difficult thing to understand what’s going on even when you actually have the animals in front of you.

Both the captive primatologists and the field primatologists can actually look at their subjects. They don’t need a time machine. Yet they often arrive at wildly different answers when explaining primate behavior. What one needs to do in general is to bring in as much different type of evidence as one can. You need different methods of watching animals. You do experiments, you create new situations to put the animals in. You expand the range of animals you study. You can also run models.

We use all these approaches to constrain our understanding. We have all these possible views we could have, so we try to constrain our explanations using the empirical data and models and so on.

Of course, to understand A as well as B, you also need to look at what humans are doing and ask: What is happening in human culture today?

So with all that said: how did humans get from A to B?

With caveats, by now we have a more constrained view of great apes living today, of bonobos and chimpanzees and orangutans and gorillas, and what their cultures look like.

And their cultures are minimal cultures, whereas human culture in many domains, especially the language domain, is very different.

There are many factors to get human language off the ground, but one of the necessary ones is the ability to copy from others, and copy how things are communicated. We learn this from the culture around us, from the usage of words and signs by others.

And it’s not just in language that we need copying abilities. We learn this way in other domains, such as dancing, in rituals and in some technology, too.

But we find this sort of copying very rarely in the animal kingdom at large. And more importantly, when we look at the great apes, we find they don’t do it very well, if at all.

As humans and great apes split from a common ancestor, we have to explain this difference. It’s possible great apes had this capability and lost it, but I’m not aware of any good theories that would explain that very well. It makes more sense to assume they never had good copying in the first place.

And the bigger question evolutionarily is, why did we humans start to do copy so well and so often, and when did this happen? This is the realm I’m working in.

Merlin Donald offers a very coherent story about going from A to B. He argues that first humans learned to mimic behavior physically, what he calls mimetic representations. Then we moved into oral language, which led to mythic representations, and then much later written language, which led to symbolic representations.

He developed this theory 40 years ago. How does this story of A to B align to your view today?

I can give a big hat tip to Donald here. He was right that copying is one of the major things. Further, he likewise argued that this copying phase, or mimetic phase as he called it, arrived after humans diverged from the great apes, and not before. The data has completely borne that out.

When it comes to the details, our stories might differ. We should factor in that 40 years ago when he wrote this, the timelines were a little bit different than they are now — archeology and paelo-anthropology is getting better, they dig more and find more.

For example, we used to assume that ‘complex stone tools’ that human ancestors were using millions of years ago must have been cultural, spreading through copying. But this is actually an empirical question. My approach was to ask, did these early stone tools really require copying? Is copying the best explanation?

And what I noticed, as others before me, is that the earliest stone tools essentially stayed the same for hundreds of thousands of years, and they look very similar across thousands of kilometers of distance, and they even stay similiar across presumable environmental changes and even across hominin species.

This doesn’t look like human culture — it doesn’t look like copying is the motor here. Like at all. Just look at how human languages change. So, our assumptions may simply have been wrong. Copying was not doing the heavy lifting that we thought it did.

So what is the storyline from A to B? My full explanatory story is not completely developed yet. We are running models right now, we have got a couple of ideas drawing from biology, especially from evolutionary biology. So there are possibilities, but I wouldn’t want to lean myself out of the window too much just yet.

That’s one reason I took up archaeology. And the timeline I find differs from Donald’s. While I cannot yet give you answers as to the why or how we moved from A to B, it looks like we began to move circa 800,000 years ago.

This relates to your theory of the zone of latent solutions for tool making. You are saying our hominin ancestors probably figured out how to make these tools independently, the tools aren’t that complex. They didn’t require copying.

I’ve written with Andrew Buskell, a philosopher of cultural evolution, about this. And essentially we say to have copying, you need three things together.

First, you need a resemblance between what someone is demonstrating — let’s call them the demonstrator — and what someone else, let’s call them the learner, is doing. That’s recurrence. This match need not be perfect, but it needs to exist.

Second, there needs to be a causal relationship between the information on how the demonstrator acts on the world, and what the learner takes in, and the behavior or output that results. The learner should causally match the output. And this requirement comes with an interesting twist — we should ideally run experiments and do interventions. If there’s indeed copying, then when a demonstrator changes the input for the learner, the latter should change their output in an according manner. The causality needs not be 100%, but it needs to be there.

Third, the learner needs to do something that they otherwise would not be doing. Not just doing something different, but doing something that goes beyond what they normally do, or could develop on their own, even if they were very motivated. This third factor is where most existing experiments fail. Many ape experiments test ape learners after giving them tasks that were within their existing capabilities, thus failing to test for this third factor.

I’ve done experiments that take into account all three factors. I have trained apes to perform forms of behavior or tasks that they wouldn’t perform on their own. It’s painstaking work, it takes a long time, but these studies can be done. I have even trained apes to be the demonstrators in some cases, much like training a dog. The results stayed the same. The apes do not copy.

Why? Usually, they fail to match the third factor.

This is always surprising to people. I wish everyone could run these experiments themselves, as it is impressive to see failures to copy. Once, a journalist attended an experiment I led. I had trained one chimpanzee to perform a type of behavior, something resembling a prayer gesture, to get a banana.

Chimpanzees do not develop this gesture on their own, but every chimpanzee wants bananas, this much is true. So the journalist thought if you bring other chimpanzees around, and they see that praying for bananas immediately gets you bananas, they will promptly pray. It wasn’t super complex what needed to be done here.

But they didn’t copy the praying behavior at all, and the journalist was just completely flabbergasted. “Why don’t they copy him?” he asked me.

That is the question indeed.

Because copying is trivially easy for humans.

For us, we don’t even think about it. When you want to make a class of kindergarteners happy, you go in and play silly imitation games. They love them, and they are very good at playing them from a young age.

But back to primate copying. The importance of the third condition is that you only will have true copying if an animal does something truly different. Suppose I train a demonstrator chimpanzee to hit a box with their hand.

Hand-hitting is an activity that chimpanzees can develop spontaneously, and that’s important. If others later do this, too, then it would not be a good explanation to say that they copied the demonstrator. They might have done so anyway, or they may just be triggered.

So, tests of this type do not test for copying. Instead, they test for whether triggerable behavior is triggerable in apes. But we already know, from many studies, that they are. Latent behavior, or what I have sometimes called latent solutions, can fulfil the first, and sometimes even the second criteria for copying. Yet they fail the third.

I’m not convinced that apes ever have shown they copy in the way that humans do. The only apes who have clearly copied are apes that were humanized – that were trained not to perform a single behavior, but trained to copy by humans. These are unrepresentative for wild apes, for obvious reasons.

There’s the famous opening scene in 2001: A Space Odyssey, and I’ve always interpreted it as the seminal moment when our human ancestors first discovered tool use. But under your theory of the zone of latent solutions, it’s likely that other hominins could have discovered the same thing independently, right?

That’s right. So I love that scene, it’s an amazing scene. I have a little glass black monolith above the door of my office, and a little Playmobile ape right next to it, touching it.

But another idea was that tool-making is what makes humans unique. Darwin wrote about apes making tools. And Jane Goodall, what was she famous for? I think most in the know would say living with primates in the wild and discovering that they make tools. She was aware she was not the first to uncover this. She had read Wolfgang Köhler’s works where he described this with in chimpanzees. She sent a famous telegram to Richard Leakey, and he came back with “oh now we have to change the definition of man,” meaning humans.

Coming back to your question, yes. If captive or wild apes can spontaneously do something, then it needs no copying. And both can make tools.

And she was featured in that the infamous Far Side cartoon. I always loved that she had a sense of humor about it.

You’ve shared a sophisticated story about apes and humans. Earlier, I felt like you were saying language is the key differentiator – is that true? Even though you’re an expert in tool evolution, is it less about tools than we might imagine?

So for a long time, I was very much focused on tools. But I have become interested in language. And I’ve contributed some work in this area. Just recently, I co-published a critique of a claim that bonobos compose their calls into new meanings, and I am co-author on another paper that regards ape communication development as absent copying.

Language requires copying. That’s not a new thought, others have said this too. And copying is not the only thing that is required here. But what it also means is that if some other activity or outcome depends on language, then it logically it also requires copying.

I suspect that this copying-need-by-proxy applies to lots of human technology and even certain types of skills. Let’s imagine carpenters — there’s individual bodily skill, but there’s also a lot you can write down, the sequence of how to put together parts, complex ideas about certain techniques. Look at the amazing carpentry in Japan, it’s mind blowing what can be held together without glue, a completely different tradition of doing things from the one I myself grew up in.

I suspect you often need language to transmit these sorts of thing from one generation to the next.

It seems to me there are those who dwell in biological evolution, and then those who dwell in the cultural evolution, and I’ve tried to merge the two to see if we can provide the full arc of the human story.1

That is much needed work, because we humans are quite bit different today than say 10,000 years ago, with essentially the same brain, biologically speaking. So, one direction is surely a better connection via neuroscience. Yet, I’m not a neuroscientist myself, and I will not become one, either — I retrained as an archeologist already. And before that I retrained as a comparative psychologist. So that’s it for me, I think. But I plan to collaborate more with neuroscientists on this question.

We need so many different fields, because the causal story is clearly complicated. It will involve interactions with the evolving culture itself and brains and anatomy and the biology underlying it, and interactions with the environment. Plus interactions across these levels. So that story is the biggest challenge: How and why did we get from A to B?

You’ve thought a lot about how humans have become humans. You’ve thought a lot about apes and their cognition, and much of that is looking backwards in time. Now artificial intelligence has come into the world, and everybody’s trying to make sense of it, and we’re looking into an uncertain future.

How do you think about AI?

That we are now essentially producing culture machines. It’s not a perfect term, but what is being created at the moment are compressed, distilled versions of our culture, essentially.

The processes are fairly well understood, of course, but the outputs have become very tricky to understand. Though I wish to add that I am in no way an expert of this topic. Yet, of course, I feel the repercussions, for example as a university teacher.

From one viewpoint, they’re massive copiers. But I’ve been thinking about what it means to be an organism and to have a culture, and I’m not sure where they fit within those.

There are two perspectives here, the scientific perspective and the ethical perspective — what may be unforeseeable consequences and the like. I’m just going to treat the academic question, but one has to be aware of both.

The models are not producing complete rubbish anymore. In my area of expertise, they can still produce bad answers, but they clearly have become less bad. Soon, I will have to stop assigning written texts to my students even in this hyperspecialized field. So there is a certain progression here.

It is much easier if the models can produce their own training data. I think AlphaGo did this by playing against itself. To the extent that you can have problems that you can represent completely in virtual space, the learning can be quite thorough, meaning that the outcomes can be valid.

But when it comes to data that people feed in, human-generated data and ideas, I am still disappointed by the outcomes whenever I check. So far, I am not seeing outcomes on par with what human researchers can produce. Instead, I see mainly regurgitations of assumptions and conclusions, regardless of their actual empirical fit.

In short, they suffer from many biases, just like humans do. Maybe this will change, and I actually expect this to change, at least somewhat.

Back to your question. I am not very interested in whether or not they will become organismic in all senses of the word. I am more interested in whether they can be evolved (or can evolve) to act as real scientists, that is, by counteracting their biases. If they become able to do this well, even when they apply themselves to messy data and messy literatures, I think they could become a game changer.

My thanks to Claudio Tennie for sharing his insights, and for our culturally evolving friendship.

Subscribe now

This is my fourth report on the reality of life at Alpha School, and it features the voices of former Alpha School students.

Over the past week, I’ve spoken to several student sources on background (for quotation without direct attribution), spanning from fourth grade up to high school, and across several Alpha School “campuses.” Every student source spent at least one full academic year attending Alpha School, and several spent multiple years. All interviews were conducted pursuant to my standing policy on interviewing minor sources.

Importantly, every student source quoted in this report attended Alpha School from the academic year commencing in Fall 2022 or thereafter. That matters because I have reason to believe that Alpha School was engulfed in crisis in the spring of 2022, and it was in the wake of this turmoil that Joe Liemandt took control of Alpha School’s operations. (My investigation into this is ongoing.)

Just as importantly, readers should know that multiple adult and student sources have described a culture of constant change at Alpha School. One adult source described Alpha School valuing what they called “spikey-ness,” reflecting a culture of rapid iteration and continued experimentation to determine what “spikes.” Further, because Alpha School is spread out across many “campuses,” practices at one location may not be employed to the same degree (if at all) at another. And the terminology employed may vary as well.

In short, it is not easy to build a mental model of Alpha School. Yet there are clear throughlines running across what student sources experienced and what they described to me, as you will see.

There is again a great deal of ground to cover, and for busy readers, I’ve again highlighted and bolded the quotes I found most notable. This report is organized as follows:

1. General experience of attending Alpha School

2. Offshore monitoring of students for “anti-patterns”

3. Rocket Ship

4. Pirate Ship (and Omega)

5. “You’ve been nuked”

6. Do students at Alpha School read books?

7. Student views of Joe Liemandt and MacKenzie Price

8. What former students would say to a prospective Alpha School family

9. What does it mean to be human?

10. Continued official silence from Alpha School

11. My brief commentary

(There’s also bonus materal that follows this reporting and commentary, including my newly activated Ko-fi account.)

Let’s turn now to what former students shared with me about their experiences attending Alpha School.

General experience of attending Alpha School

In my prior reporting, adult sources described Alpha School’s culture in stark and concerning terms. My student sources echoed these descriptions of Alpha’s culture, and provided greater context on their respective experiences.

Going to Alpha was not good. It was rough, very rough. There was two months where I cried every single day.

It was very harmful to my mental health at the time. I was so distressed by it that my mental health was going downhill real fast.

After the first week [of high school], which was fun, everything changed. It was like lockdown. It all started to change, like the science experiment [had] really started.

The best way to describe the vibe was dystopian. Computers, technology, very futuristic, but also very distant and cold.

It is notable that multiple adult and student sources have used the word “dystopian” to describe Alpha School.

Offshore monitoring of students for “anti-patterns”

In March 2026, MacKenzie Price appeared on a podcast called Class Disrupted, hosted by Diane Tavenner and Michael Horn (I have also appeared on this podcast). In the interview, Price stated, “We have an anti-pattern, which is basically an improper use of engaging with the apps.”

My student sources confirmed this is true, while providing important context on how Alpha School detects “anti-patterns.”

They hired this data team of people that didn’t live in the U.S. and they would watch us through our computers and it felt really sketchy and scary. It felt like it was more of a science experiment, and I was more of a lab rat rather than [having] someone that actually cared about my learning.

I was getting really frustrated about that throughout the year. Every time they would change something, I would be like, why are we changing this?

We couldn’t tape over the computer or anything like that. They would get really frustrated if we taped over it. We had to keep the camera on.

[The software] would also track your facial expressions, and be like, “This is the emotion that you’re evoking while you’re talking about this.” And every single time, it would tell me that I look really sad, which is fair because I was really sad, because I was crying every single day.

They would make me and a few other kids take standardized math tests pretty much every day, and then any question that we got wrong, we would have to get on a Zoom call with some random guy from normally India and explain to him every single thing that we got wrong and why. And I hated that.

I was a guinea pig, and I knew that I was a guinea pig. They were not subtle about it.

Anti-patterns were like this AI system that decided whether or not you were doing good enough. It saw that your eyes were moving away from the screen, you’ll get flagged for those.

You did not want to get an anti-pattern. It was creepy.

Also on anti-patterns, you could just jiggle the mouse repeatedly, and it would read that as you “thinking,” and so I could just…wiggle the mouse with my hand.

Whether students at Alpha School should be considered part of the company’s software product development team is an open question, and so too is whether this raises issues of child labor law violations or other forms of civil liability. I am referring potential claimants against Alpha School to the EdTech Law Center, a firm that pursues tech-accountability litigation against Big Tech and other bad actors.

As flagged in my initial special report, Alpha School has engaged Carl Hendrick as consultant to its software development, and Hendrick has stated publicly that “in the future, this will mean cameras on, and an AI monitoring student behavior, their latency, what websites they are looking at, their ability to focus and then producing a report for a human tutor.” He further predicted, “you’ll see articles about ‘Orwellian spyware.’”

Based on what my student sources told me, it appears the future Hendrick describes is already present at Alpha School, at least to some degree. As for his prediction Alpha School’s activities would be described as “Orwellian spyware,” given that 1984 features a world wherein The Party uses omnipresent screens to monitor the activities and emotions of inhabitants of Oceania, the analogy appears apt.

I also reached out to the Class Disrupted co-hosts Michael Horn and Diane Tavenner, who I’ve both known for years, to comment on how student sources described the offshore monitoring of their behavior and the detection of so-called “anti-patterns.” I have not heard back from them as of September 23, 2026 when this story was first published.

I will update this report if they choose to respond.

Rocket Ship

In October 2025, MacKenzie Price authored a blog post on the official Alpha School website titled “4 (Controversial) Motivation Systems That Actually Work.” In her post, Price identifies something at Alpha School called “Rocket Ship,” which she describes as an “internal program that offers the highest level of autonomy to kids.” She further states, “I’ll be honest: in Alpha High’s Rocket Ship program, you barely have to come to school.”

My student sources confirmed this is true, while providing important additional context on student life in the “Rocket Ship.”

I was in the what they called the Rocket Ship, which was basically their smart kids club. If you were really good in school, if you got great accuracies, completed everything you were supposed to do consistently, you would get put into the Rocket Ship.

Really all it was was a group of students they thought were elite, and better than the rest of the students. They had more privileges. They had really no rules. They could do whatever they wanted. They didn’t really have to come to school during the school day. They could choose to do work when they wanted. They were given like snacks for free. All these other privileges that were completely separate from the [non-Rocket Ship] students.

They had certain rooms they could go in without asking. We could not go in them. The regular students could not go in them. So they pretty much segregated these higher students, and they had more privileges.

They picked a couple of the best students, separated them from everybody else, and we got to work in a special room, listen to music, eat snacks, go to parties. It was really fun for me because I was in [the] Rocket Ship… but looking at it from outside in, [it was] horrible for everybody else, having to watch as everybody else threw parties and ate snacks and went on field trips.

They [had] kindergartners in it, and it feels so weird. It seems so sad watching these kids trying super hard, and then they don’t get their XP [Alpha’s 2-Hour Learning metric], and they don’t get in the Rocket Ship.

I saw this kid trying to read [in the app], and then he just stormed off crying. They have them moved to another room. I felt bad for the kid, and it seemed dystopian. It felt like weird.

I agree with MacKenzie Price that this motivational system is indeed controversial.

Pirate Ship (and Omega)

Multiple student sources described another internal motivation program at Alpha School, referred to in the lower grade levels as the “Pirate Ship.” Unlike the Rocket Ship, I could not find any reference to the Pirate Ship on Alpha School’s website or anywhere else publicly.

Here is how my student sources described the Pirate Ship.

I was in the Pirate Ship. We got kicked down grade levels, and we had to work all the way back up. We were completely shut off from the rest of the students. We had completely different privileges. We weren’t able to do any of the workshops or anything that they were doing in eighth grade. It’s a morbid term but…[it was] pretty much a concentration camp of students. Until they were up to the standards of what Alpha wanted us to be.

Some people could never even finish the list [of Pirate Ship requirements] because some of it was unbelievably difficult. We had to type at 120 words per minute or something like that. Some people can’t type that fast.

The Pirate Ship, it felt like you were bad. It’s like you were being labeled as someone’s idiot.

Some students never left the Pirate Ship.

The moment that they stop caring about you, life becomes shit because they’re gonna put you in the pirate thing, and then you’re just stuck there all day.

Basically, all of the students that didn’t complete boot camp within the allocated time period would get put into the Pirate Ship. They would have to do extra work all the time, and they wouldn’t get to have the clubs that [were part of the rocket ship]. And those kids would just dedicate all of that time to getting back out of the pirate ship.

They would be confined to one room during club times, and not get to participate in any of it.

In addition to the Pirate Ship, one student source described a similar internal program within Alpha School’s high school as something they called “Omega.” As with Pirate Ship, I found no reference on Alpha School’s website or anywhere else that publicly acknowledges this program.

Here is how “Omega” was described to me by this student source.

We [were] just completely separated, and it felt like a hierarchical system. If you’re liked by the school, you’re in the Rocket Ship. If you’re not, you’re in Omega.

They kicked me down to sixth grade math when I was in ninth grade. They screwed my credits up. They’re doing it deliberately so people cannot leave. If your credits are messed up and you try to go somewhere else, you might not be able to graduate.

There was a handful of us that they just kind of shoved in a corner. Like that’s why we were in the Omegas because they didn’t like how we figured out that we weren’t actually learning anything. We were just kind of robots.

It’s not marketed as “you’re bad.” But it was just like, “you’re leaving.”

Whether credits are being manipulated by Alpha School leadership to prevent students from leaving is one of many things I’ll be investigating further. Also, based on conversations with both adult and student sources, I believe there may be significant hidden student attrition at Alpha School and/or segregation of academically underperforming students from their top-line metrics — my investigation into this issue is ongoing.

I also find it noteworthy that one student source described students being stuck in the Pirate Ship following boot camp. This hints at a possible consequence that Alpha School’s freshmen class may soon face if they fail to complete their “life skills” in the eight-week boot camp they are currently participating in.

“You’ve been nuked”

In my prior reporting, multiple adult sources expressed frustration regarding the glitchy performance of the TimeBack/2 Hour Learning software. My student sources echoed this complaint, and several mentioned something they called “getting nuked.”

Here is how they described this phenomena.

The tracking software would mess up regularly, and there was not really a way to fix it. They would block my computer. When you didn’t have the right like hours or accuracy in your previous day’s work, they would do something called nuking.

Basically it would shut down your entire computer [other than the 2-Hour Learning app], and just show a jokey screen of a mushroom cloud. “You’ve been nuked.” You’re doing something you’re not supposed to on the computer right now because you didn’t get yesterday’s accuracy high enough.

[Adults at Alpha School] would have to communicate with the people who ran the software side of things, who were not in the building. A lot of them were in India, outsourced to other countries, and it would take forever for them to get back to us, and it would make it really hard to actually complete my work.

I remember getting nuked, and I remember all my friends getting nuked.

You were stuck in the apps and [telling guides] “no, I finished it,” and they would be like, “nope, now you got to go redo it again.” And the guides could not actually override [being nuked], they had to contact someone else, and sometimes that would take a whole day. It was a very convoluted process.

And they would change something [in the app], and it would take them a week to get anything right. So everything was screwed up. I didn’t feel like I was actually having measurable learning.

Whenever somebody messed up on something or flagged, this big screen [would] pop up [stating] “you can’t access this anymore.”

[Q: Did it show a mushroom cloud?] Yeah, and a raccoon.

I found no public reference by Alpha School leadership, or anyone else affiliated with Alpha School, related to students “getting nuked” when using Alpha School’s software.

Do students at Alpha School read books?

In my prior reporting, adult sources suggested Alpha School places little emphasis on literature. So I asked former students if they read books when they attended Alpha School.

Here is what they told me.

Not a single one. I had one English class where we were told to read a book. No one did. I just read the Spark Notes.

Not full books. We read some poems, because I asked to.

We never had a curriculum of books to read. Never a full book, only passages [on the software app].

I plan to investigate further whether or how Alpha School covers other core academic subjects beyond math and English, such as history, social studies, or the humanities more broadly.

I would encourage students at Alpha School, or any student for that matter, to read 1984. I independently read this novel as part of my AP English class in high school, and will never forget my relief seeing the first essay question posed (paraphrasing from memory): “Describe a character who bears great influence throughout a novel but never appears.”

I wrote about Big Brother.

Student views of Joe Liemandt and MacKenzie Price

Every adult source I’ve spoken to has expressed a strong opinion about Joe Liemandt and/or MacKenzie Price, Alpha School’s leaders, and that same is true for my student sources. Here, I share student commentary on Liemandt first before turning to comments on Price.

Joe Liemandt describes himself as the “Principal at Alpha School.” Typically, a school principal is someone onsite at a school campus with ultimate responsibility for overseeing the school’s day-to-day operations, at least principals of “normal schools.”

Students described Joe Liemandt’s role at Alpha School very differently.

He didn’t care about the education. He cared about…creating businesses that he could be part of.

It seemed like he was more after the ideas of the students, profit-wise, than the actual learning the students had.

[The guides] had very little power. Their entire job was to enforce what Joe Liemandt wanted them to do. Alpha sort of markets itself as being student run, where you have control over your own learning. It is not student run. It is exclusively run by Joe Liemandt.

There was an illusion of freedom, but you only had freedom within what he wanted you to do. He wanted us to become entrepreneurs. His money is his power, and he’s not afraid to use it.

Nobody could tell him what he couldn’t do, because there’s no repercussions for him. He just fires everybody that doesn’t like him. If a student is going against his wishes, he can just get rid of them. What he is doing is hurting people, [but] if they speak out, he can just kind of get rid of them.

He’s not disclosing to students that when they enter into Alpha, that it is an experiment. He’s saying this is a school and it’s going to work. It’s not a school and it’s not guaranteed to work. What it is…is a science experiment that’s been framed as a school.

If Joe Liemandt is not functioning as the actual school principal of Alpha School as that role is conventionally understood, I was curious whether MacKenzie Price might be.

My student sources did not think so. They instead described her role differently.

The only reason she visits the campus is social media. She’s always doing the social media, and convincing parents…There’s a sense that she has hierarchy over you.

She was just kind of Joe’s minion. She’s sort of framed to be at the same level as him, also in charge. But I never heard about anything that she actually did. Nobody knew if she was doing anything.

I do not like Mackenzie Price…She’s just talking about using AI for learning. They don’t use AI. We used Excel and apps. That’s not AI.

I’d see her sometimes, but it was all filming and speaking with the Rocket Ship students and putting them on camera. I never had a conversation with her. She would focus on the students that were elites, because that’s what they wanted to portray.

She came to the school to film a video or something. We never got to meet her. She was always kind of distant…I don’t want to put words in her mouth and say she saw us as “less,” but she never gave us the time of day.

[Q: Did she ever say hello to you when visiting?]

No.

What former students would say to a prospective Alpha School family

I asked several of my student sources what they wanted to say to prospective parents or students who might be considering enrolling in Alpha School. This is what they told me.

It is reasonable to want to know things, and it is reasonable to want to prove things. However, it is not reasonable to treat students and schooling like an experiment without any regards for how students actually feel about it [or] without regards to the ethics of the situation.

You’ll regret it for the rest of your high school career, maybe your college career, because your credits will be ruined. You will feel like you’re lost.

While I was there, I was not happy.

It’s gonna be cool the first few months and then it just becomes shit.

Just don’t. Do better.

What does it mean to be human?

Earlier this year, Teddy Rosenbluth of The New York Times reported on my father’s passing last year and the role of AI played in hastening his death. She wrote, kindly and accurately, that “the rise of generative AI changed his [my] view of the technology, though. It offered a window into many of the questions [Ben] had devoted much of his career to: What makes us human?”

Unexpectedly, several student sources shared with me that they too are grappling with this very question. And so I asked them to share their still-emerging perspectives on what it means to be human.

Being human is about learning and growing and expanding who you are. Being a functional human being is not about your SAT score.

To have love. To see a world that can be connected if we put enough passion into it. Passion fuels everything.

To be human, you have to have connections, and you have to be honest. And to be human, you have to love.

In being more human, my biggest one is contribution. Being able to contribute to the world, a huge way of doing that is through contributing to the next generations of people.

The last student quoted above told me they are considering becoming a teacher. A human teacher.

Continued official silence from Alpha School

On Friday September 18, I emailed the following to Anna Davlantes, Alpha School’s Chief Communications Officer, to alert her to my forthcoming reporting and to invite Alpha School to respond:

I received no reply.

My brief commentary

My ongoing investigation

Supporting Cognitive Resonance: donations now accepted

In the wake of my reporting, I’ve received a large number of new followers (welcome!) and a non-trivial number of pledged subscriptions to this newsletter — always deeply appreciated, thank you. I have not activated this feature on Substack for many reasons, but it does feel very validating.

That said, and as I mentioned in my third report, no one is paying me to do investigative citizen journalism, and I’m starting to incur some non-trivial expenses related to it, such as improved IT security and home security measures, expanded professional insurance, dedicated phone line for reporting, etc. It’s also crowding out my time for undertaking other revenue-generating projects, no small thing for those of us who run our own “micro think tanks.”

As such, I’ve established an account with Ko-fi to accept financial contributions to support Cognitive Resonance. Any support you choose to provide will be greatly appreciated.

You can donate by clicking here: Support Cognitive Resonance

Bonus coverage

What’s that, you say you want even more Alpha-related content? Well then I have good news!

First, I appeared on the Kill the Computer podcast last week to discuss my reporting on Alpha School. The hosts June and Caleb did their homework and asked great questions, and our discussion covered a wide range of Alpha School’s activities. The first 30 minutes are freely available, and the entire episode is accessible via Patreon.

Second, a few weeks ago my friend and veteran educator Larry Ferlazzo asked me to write a high-level overview of Alpha School for his standing “Classroom Q&A” column with Education Week. Believe it or not, the majority of this essay was drafted prior to my investigative reporting of Alpha School.

And for those wondering, there is a non-Alpha-related essay coming later this week. I promise.

Subscribe now

Let’s try to nip some odious stupidity in the bud, shall we?

When I published my special report on Alpha School boot camp last week, I knew that it would cause a frenzy of activity among its terminally online supporters. These people, yeesh. Many appear to have no official relationship with Alpha, but they are no doubt attracted to its public posturing as “the future of AI-driven education” or whatever, and maybe more importantly, its repeated denigration of so-called normal school. They have embraced an ethos of intimidation and gaslighting in failing pursuit of assassinating my character.

I call them the Alpha-Betas.

There’s no time to marshal arguments against their dystopian vision — so much investigating to do! — so instead, here’s a book recommendation (which of course they won’t read because books terrify them): Democracy’s Schools, by Johann Neem. It’s a lovely history that tells a comprehensive story that too few Americans appreciate, namely, that democracy and public education, emphasis on public, are bound together. It is no coincidence they are eroding in tandem too. My friend Kevin Carey echoes this point here.

But back to the Alpha-Betas. It took them a few days to figure out how they wanted to commit mass ad hominem fallacies, but eventually they coalesced around repeatedly suggesting I’m “creepy” because I want to “DM teens.” They want people to wonder if I’m a pedophile.

Leaving aside how disgusting this is, as a rhetorical move I find this fascinating. Since I know myself and they do not, I know that I am none of the terrible things they repeatedly and grotesquely hint at. Unfortunately, however, I have no doubt there are many people on Twitter who are dangerous, and very much would like to DM with teens for nefarious reasons.

This is exactly why Alpha School should not force its students to be on social media, you jackasses.

Obviously. Obviously! In my special report, I quoted two veteran educators who observed that in the US and England, this would be immediate grounds for termination in the public-school system. The world is filled with people both good and evil, and social media attracts the latter because of its cloak of anonymity, so forcing children to use it is wrong. Stop doing it, Alpha School! Stop it.

There is one active online presence, identified only as Wendy with the Twitter handle @teachthemx3, who has been particularly keen to suggest I’m a creepster. Unlike the Alpha-Betas generally, however, Wendy’s profile indicates she presently works at Alpha School:

Because of this, I asked her to reconcile the glaringly obvious contradiction between her (wrongly) believing that I pose a danger to students on Twitter with Alpha School’s official policy of requiring students to use Twitter. Took me a few tries, but eventually she said this:

Leaving aside the still-open question of whether parents or students actually knew anything about boot camp activities when they enrolled — and I am deeply skeptical, as no parent I’ve spoken to had ever heard of it prior to my reporting — if Wendy or anyone else is concerned with adults DMing minors, their issue is with Alpha School leadership and its official policy, not me. They are putting minors at continued risk online. They require students to obtain verified Twitter accounts and gain 100 followers and tweet every day. This was Alpha School leadership’s decision. They own it.

What’s happening right now within Alpha School is massive cognitive dissonance, and Wendy is Exhibit A. Alpha School has deliberately cultivated a culture that depends on triumphal and relentless affirmation of its superiority. Unexpectedly, it is now being investigated by someone with 20 years of experience in education, extensive knowledge of cognitive science, AI and ed-tech, an entrepreneur who also happens to be a former Deputy Attorney General, and someone who simply does not believe the mythology Alpha School has cultivated, or the claims they make about their success.

So, what’s the actual basis of the Alpha-Beta smears? I’m sorry I have to waste anyone’s time with this (including my own), but I will now respond to the pathetic arguments they’ve employed.

The main line of attack is again that I want to DM students. Nope, I do not. What is true is that on September 3, the very first day I discovered the student tweets about boot camp, I posed questions to two student accounts publicly and said that they could DM me if they preferred to correspond that way. This included the student who had tweeted to MacKenzie Price about their concerns regarding the clash of morals surrounding the workshops — I sensed a young conscience in crisis. In the moment, I had no idea I was about to pivot my life toward investigating Alpha School, but it was beginning to dawn on me that something terrible might be happening to these students, and they might be afraid to respond via public tweet.

That’s it, that’s literally all there is to it.

What’s more, if there was any question of my motives at the time, my reporting has now answered it. I am obviously reporting on student life at Alpha School, and I want to hear from former and current students to learn more their day-to-day experiences. Alpha School may prohibit its adult “guides” from talking to children so as to not “compromise its model,” but out here in the Land of the Normal, adults and kids can have conversations. It’s fundamental to being human, actually.

In any event, this is now a moot issue. My story began with the students in boot camp, but my interests in Alpha School have broadened far more widely. And now that the students at Alpha School know of my ongoing investigative efforts (and they know), Twitter is no longer a fruitful or necessary medium of communication — and thank goodness for that, because it’s so very awful. Instead, if a student wants to talk with me, they should do so pursuant to my policy about communicating with minors that’s posted here.

But wait, there’s more inanity. It seems all-but impossible that the remaining Alpha-Beta character attacks could rely on even weaker “evidence,” but then, I’ve observed they aren’t the sharpest tools in the shed. The entirety of their remaining argument is that because I wrote that I “kind of adore” the students in boot camp, and joked about getting a drink with one when they turn 21, I must be the next Jeffrey Dahmer.

My first response: You have got to be fucking kidding me. My second response: This line of attack is actually quite revealing.

Let’s start with adoration. When I say I adore something, I am expressing fondness towards that something, not desire. Here are a few other things I adore: my 98-year old Aunt Bev and her amazing stories; the little rabbit that hops in my yard in the early morning hours that I’ve named Mortimer; Alfalfa the cat who likes to sleep under my truck who I’m worried will one day eat Mortimer; the joyous chants of the kids in my kickboxing gym; Steve Raible yelling “Holy Catfish!” when my beloved Seahawks make a great play (which is often); and Dolly Parton (RIP).

Any rational human being who possesses the virtue of empathy would understand that when I wrote I “kind of adore” Alpha students, it is in this spirit. But of course this virtue is in short supply for many — most? all? — of the Alpha-Betas. And then there’s Twitter itself, flailing under the leadership of its abominable CEO, amplifying sociopathy at scale. Yet another reason why Alpha School should not force children to use it.

Finally, as to my obvious joke about hypothetically having a drink with a student in seven years time, here’s some bad news for the Alpha-Beta kill-joys — if one day in the distant future someone who attended Alpha School wants to grab a beer with me (again, after they turn 21), they will be free to do so, because humans have agency. I suspect attending Alpha School will be the source of many conversations throughout their lives, the same way people are often interested in my Enron stories.

Human curiosity, I adore that too.

I want to turn now to something that makes me sad, but also angry.

Carl Hendrick and I have known each other a long time, as we’ve both been part of a shared effort to spread ideas from cognitive science into education. I’ve long considered him a friend. More than that, actually — as I said to him via email earlier this week, for many years I considered him my “intellectual doppleganger” on the other side of the pond.

On September 4, knowing his very publicly visible role as advisor and advocate for Alpha School, I sent him a DM via Twitter asking him to comment on my forthcoming story on boot camp. I specifically referenced the biometric monitoring and the weeklong discovery learning requirement to complete an AP requirement. He responded as follows, with my emphasis:

“I have never heard of anything like that and it doesn’t sound true but I am going to call Joe [Liemandt] as that is quite the claim.” (Alpha School has since confirmed it is absolutely true.) That evening, I forwarded Hendrick the email I’d sent to the Alpha School PR rep outlining what I’d discovered in even more detail. He replied:

Yes! Yes these are concerning. Hell, everything I’m uncovering about Alpha School is concerning, shocking, harrowing, pick your word. And I believe Carl was telling me the truth in his initial statement that he had never heard of anything like what’s been going on boot camp.

Yet, sadly but also infuriatingly, to my knowledge Carl’s never expressed publicly what he said to me privately. Instead, he has since doubled down on his defense of Alpha School on social media. I cannot fathom it. I’ve made repeated entreaties to him to just resign, but these efforts have obviously failed spectacularly, and they won’t continue any further. But I find myself wondering this:

Finally, I saw some tweets floating around suggesting I’m “threatening” to talk to The New York Times. Not true, and this would be a strange threat to make, since The New York Times reached out to me to last week, of its own volition, and I’ve already gone on the record. I’ve also talked with two other reporters in San Francisco and Chicago — it’s because one asked about the Alpha-Betas that I’ve written this essay. I’ve also just recorded a fun podcast that will drop soon, and while drafting this post I received an inquiry from NPR.

We are only in the beginning stages of this scandal. The dam is breaking and already we see hints of vibes shifting — see, for example, this blog post from the Fordham Institute, a center-right DC-based education think tank, titled “Alpha School’s Theranos problem.”

The analogy is apt.

None of what I’ve written here will matter to the Alpha-Betas, of course. My arguments won’t persuade them, in fact they won’t even make sense to them, because we live in separate cognitive worlds, with values that are an anethma to each other. There is no cognitive resonance between us.

Sucks to be them! But now I can point to this post should anyone ask why they are so deranged. And while the Alpha Betas will surely cook up new and perhaps even more disgusting accusations in the future, I will remain undeterred, and here’s why:

My late-in-life mentor, the brilliant yet deeply humble Jim March, often wrote about the story of Don Quixote as a story of finding purpose, and to embrace a life well lived not in terms of its material consequences, but rather in the pleasures found in the pursuit of duty. Quixote pursues knight-errantry and tilts at windmills because “he knows who he is.”

I know who I am. And I have found a new duty to pursue to bring truth to light.

Subscribe now

I’ve turned comments off for this one for obvious reasons. I promise there’s non-Alpha School related content coming soon.

UPDATE September 19, 2026

I’m paying no attention to the Alpha-Betas — I’ve got much bigger fish to fry at the moment — but Cognitive Resonance has engaged legal counsel to keep tabs on the more egregious defamatory statements. Go outside Alpha-Betas, there’s a big beautiful world out there to explore.

Because such sources may (and likely will) be minors under the age of 18, extra ethical sensitivity is required. Sources who are minors and feel the need to ask for permission from their parent or legal guardian before contacting me should do so. Minor sources should show them this policy as part of that process.

With the appropriate permissions, minor sources should email info@cognitiveresonance.net with the subject heading “ALPHA SCHOOL - STUDENT,” with a brief one-sentence description (preferably without the use of AI) outlining what they would like to discuss with me. It should indicate if they want a third-party present for our conversation. We will then set up a time for a phone call or video conference or other means of communication, depending on the minor source’s preference.

During any such conversation, minor sources can designate whether what they tell me is on or off the record or on background, in whole or in part. Minor sources will be reminded that anything they say on the record may be used in my reporting, and may subsequently be shared on social media.

In conducting any conversation with a minor source, I intend to follow the Interviewing Children guidelines promulgated by the Education Writers Association, and authored by Sarah Carr, Director of Columbia University’s Spencer Education Journalism Fellowship.

Should any minor source contact me using social media, I will direct them to this policy and the process it describes, and then conclude our communication. In an abundance of caution, I will not solicit information from current Alpha School students on any social media platform.

This policy is effective as of September 15, 2026 and will remain so for the duration of my investigation.

What I uncovered about Alpha School boot camp was a tiny tadpole in a much larger and very polluted pond.

People are reaching out to tell me their stories. In the wake of my special report and follow-up story last week, I’ve spoken to numerous former Alpha School parents, ex-Alpha School employees (so to speak, see footnote), current and former Alpha School contractors, and former employees of Joe Liemandt’s other business ventures.1 At least one currently enrolled student reached out to me, but later got cold feet about talking, which is more than understandable.

Because people are scared.

For many reasons, the people I’ve corresponded with have expressed worries about speaking up. I do not blame them. There’s no upside for them personally. My reporting raises the question of whether Alpha School is a cult, and however one chooses to answer that based on what we’ve learned, there’s no question that Joe Liemandt possesses the resources that strike fear in people. Cults depend upon mythology in order to maintain their appeal, and they often will strike back hard at efforts to pierce their mystique — a great deal of First Amendment law has been created via the Church of Scientology. I have been consulting with practicing attorneys regularly throughout this process — my thanks to all of you, you know who you are — which is not something I’d ever anticipated writing a newsletter about cognition and AI.

This story is massive. I have spoken to numerous reporters who are working their own angles, and their stories will drop soon. I have also initiated efforts to farm out some of what I’ve learned to other investigative journalists, because I do not have the time nor the resources to pursue every lead I’ve uncovered. My main focus at this point is the student experience at Alpha School, for reasons that will soon be apparent. But there are many issues I’m trying to hold in my head.

Some have asked me why I am doing this, and who might be paying me. The short answer to the latter question is no one, as I never expected to become thrust into the role of “citizen journalist,” this was not on my career agenda.

As to the why, the short answer is that children are at risk. As I mentioned in my initial report, it was finding that solitary student tweet sent to MacKenzie Price, the student wrestling with the morality of boot camp activities, that shifted my perspective from bemused curiosity to genuine concern. What I’ve learned since about Alpha School has shook me to my core. I’ve cried multiple times in interviews.

In what follows, I share what I’ve learned about various aspects of Alpha School activities and culture. Given the palpable fear of my sources, I agreed not to name them in this story and to remove certain identifying details. I’ve used the designation “observer” to describe former Alpha School employees, as well those who went through the hiring process but did not join Alpha School, and contractors who had full-day in-person visits. Parents are identified seperately. I’ve added highlighting for emphasis (and skim-reading, no judgments here); reporting and commentary from me are in standard text format.

There are four sections to this report:

1. Observations on TimeBack (formerly “2 Hour Learning”)

2. Observations on the Dean of Parents role at Alpha School

3. Observations on Alpha School culture generally

4. My brief commentary

Let’s dig in.

Observations on TimeBack (formerly “2 Hour Learning”)

The Alpha School model involves students using software on computers for at least two hours each day. This was and sometimes still is called “2 Hour Learning,” but in November 2025, a blog post by MacKenzie Price on Alpha School’s website suggests this activity has been rebranded as “TimeBack.” As Price explains, the name was apparently chosen because the goal of Alpha School is “to give kids their time back.” Makes sense, at least as nomenclature. The same post states, “TimeBack is a $100 million project to build an AI platform that helps students learn 10x faster than traditional classrooms.”

None of my sources suggested that students are learning 10x faster because of TimeBack. Here instead is what they told me.

Observer:

I watched kids clicking on a bunch of crap. And there is just one image that really stuck out to me.

When I was young, I went to a school where we learned about nature and went outside, and there was a fish pond where we would catch fish. But then I saw this kid looking at a screen with a picture of someone holding tadpoles in their hands, and the app was talking about tadpoles, and I was like, man, that is so pathetic. The families are paying tens of thousands of dollars per year to have their kids look at tadpoles on a screen, rather than experience it for themselves.

It’s pretty dystopian if you think about it.

Observer:

The kindergartners were breaking my heart. They were falling out of their chair. They were putting their heads down. They were clearly not enjoying the learning, at all. One student said to me, “I feel like I’m going to fall asleep in the middle of my work.”

Parent:

Guides and parents were explicitly told not to help or get involved with the kids’ academics. Part of them developing “grit” was to let them struggle. Guides only saw the academics through the data on the Dash, not what the kids were struggling with, not issues with comprehension, just whether they did it or not.

It was extremely frustrating for kids to have the adults in the room do nothing to help but [instead] redirect them back to the screen, and dangle rewards or the fear of having privileges taken away.

Why might guides be prohibited from helping students during the TimeBack period? One source said to me that Alpha School wants to test its software free of the confounding variable of human intervention. Notably, the 404 Media story on Alpha School contains a quote from a former Alpha School employee stating that, “students are being treated like guinea pigs.”

Whether Alpha School guides are formally or informally be prohibited from helping students is an open question. I reviewed one job posting for a guide role on the Crossover website and found this description noteworthy, with my emphasis2 :

• You are not a teacher. At Alpha, students learn academics through AI-powered adaptive learning apps. No lectures. No textbooks. No whiteboard. Motivation and life skills, the two things schools normally squeeze into the margins, are the whole job here.

So it appears motivation and life skills are the “whole job” of being a guide.

Here are some additional comments related to TimeBack.

Observer:

Many kids really hate TimeBack. Kids will cry in frustration with the apps, and Alpha School leadership is very dismissive about kids that don’t like it. They’ll say, “they just don't get it.” When the kids are struggling, as I understood it, the guides are not allowed to help them. They might refer them to a special tutor…but that’s done offsite over video.

Meanwhile, the guide is standing right there and doesn’t step in.

Parent:

Alpha told the guides they weren’t supposed to help the kids, but they did anyway. They’d often encourage kids to figure things out on their own. That was baked into Alpha’s DNA. [There was] a fair amount of “figure it out yourself” — independent learning was treated as a core value.

To reiterate what I’ve stated previously, Alpha School’s apparent emphasis on something akin to discovery learning is remarkable, given the prominent involvement of Dr. Carl Hendrick in its activities. Few people have been more vocal than Hendrick in criticizing this approach to instruction (or lack thereof).

I asked Dr. Paul Kirschner, co-author with Hendrick of the book How Learning Happens, to respond to the parent’s observation regarding independent learning. “Discovery learning is neither an efficient, effective, nor satisfying way to learn,” he said. “Discovery is the epistemology of the expert and not the pedagogy of the novice.”

One additional note: I have a growing sense that there may be significant differences between student experiences at Alpha School in Austin — where its history can be traced back to 2014, according to WIRED — and those of students at newer “campuses.” The parents based in Austin were uniformly positive about the guides that worked with their students, and one suggested the guides will help struggling students during the TimeBack period. This was not the case for parents elsewhere. More investigation is needed here.

Alpha School did not respond to my request for comment on whether guides may intervene if a student is struggling when using the TimeBack software.

Observations on the Dean of Parents role at Alpha School

Last week, the official Alpha School Twitter account tweeted this:

The Dean of Parents role at Alpha School is an interesting one. I reviewed one listing on the Crossover website for this position; it contains a job description that states that “you keep the model intact”:

It is perhaps noteworthy that deans are supposed to get “ultra-high-net-worth” familes to champion 2 Hour Learning rather than Alpha School itself. I also found this “key responsibility” interesting, which again states that “no parent crisis is resolved by compromising the model”:

So preserving the TimeBack model seems important, even if parents object.

My sources had more to say about the Dean of Parents role.

Parent:

I saw the job description for the first Dean of Parents when it was brought to a parent group meeting, and it was remarkable — it basically said your job is to be a buffer against “wealthy, entitled, litigious parents.” It was just talking down about the parents.

The whole Dean of Parents position is a big reason a lot of families left. Why do they even need a buffer?

Parent:

[The Dean of Parents] essentially functioned as the HR department, taking complaints, protecting the company, and maintaining order. The position was framed to us as a liaison to streamline communication between parents and the school, which was a major complaint at the time.

Initially, I felt bad for him because I believed he had bought into the Alpha narrative but [he] slowly realized the chaos he was coming into. He did significant housekeeping, created systems, reframed issues, and put out fires while Alpha was selling its charter around the US as a proven model. However, despite everything brought to his attention, the school never addressed the underlying issues. He served as a pacifier, repeating school talking points he had to know were untrue.

I hold him responsible for the ongoing harm experienced by students. Very few people within the organization had direct contact with upper management, academics, guides, and parents simultaneously, but he did.

Observer:

The dean handles interfaces with the parents. And I thought that’s odd, because as a teacher, one of the best things you can do is know your family.

I have reached out to one person via social media who appears to have recently joined Alpha School as assistant dean of parents at a “campus” in Texas. I encourage other current and former deans to contact me to share their experiences.

I asked Alpha School to elaborate on what it means for “families to be advocates for TimeBack/2 Hour Learning,” and what it means to “compromise the model.” I received no response.

Observations on Alpha School culture generally

In January 2026, Alpha School co-founder MacKenzie Price published a blog post on the Alpha School website with the title, “Your Kid’s Social Life Matters More Than Academics.” In that post, Price describes various aspects of Alpha School culture, including it being a place where “being a ‘try-hard’ is cool.”

Similarly, one parent with a student currently enrolled in Alpha School’s program, a vocal proponent of Alpha School on social media, echoed this sentiment in mid-August3:

I have little doubt that Alpha School makes business-like activity a central focus, as discussed in my first report in relation to boot camp. But my sources describe a somewhat different culture at Alpha School.

Parent:

It’s like Y Combinator for high schoolers, or a Shark Tank knockoff. It’s ridiculous. The kids are being done a huge disservice — none of it will matter once they’re out in the real world.

Observer:

The culture is to cultivate Silicon Valley startup founders. It’s not to create a well-rounded human.

I think of education as cultivation of the person. But there, if the kid is not interested in what is being taught, they don’t have to learn it — unless it’s teaching to the test with TimeBack, or getting them to crush the SATs.

I don’t think students have to read any literature or great books.

Observer:

Kids are very good at mimicking adults. They’re mimicking some idealized version of the billionaire who started his thing in a garage and grew it into a billion-dollar company, and that sort of ideal that has permeated through the culture.

Parent:

Alpha is Silicon Valley’s fake it-until-you-make-it culture taken to an exponential level. They’ve threatened people, claiming violation of an NDA. They use children as a shield, a buffer against criticism. They want people talking about the kids as a shield against criticism of the school. What they really care about is selling software. They’re faking it until they make it — pure charisma, gaslighting, and intimidation to push the model. It has a cult-like atmosphere.

Observer:

I started reading Punished by Rewards by Alfie Kohn, and I think Alpha School students are being punished by rewards. They’ve set up a system where you do things to get something tangible. Education is should be a reward on its own — I’m not against stickers or things like that — but overall, it’s the relationships and discussions and conversations that enrich a classroom. But at Alpha, students are purchasing smartwatches with their alpha points.

Alpha School did not respond to my request for comment on the highlighted observations shared here.

My brief commentary

I was so very proud of my daughter when I described the events of Alpha School boot camp and she replied, without hesitation, that’s not okay. But now I’ve learned so much more about Alpha School, and thus I can state plainly, drawing upon my 20 years of experience working in education as well as my basic sense of human decency:

What’s happening at Alpha School is immoral.

And I’m planning on staying on this beat until the full truth comes to light.

Subscribe now

My friend and moral conscience Audrey Watters offered her thoughts on Alpha School over the weekend in an essay titled Click Here to Comply. Grateful for everything you do, Audrey. You help us find the light.

And The Simpsons helps us find the levity, as least in the good seasons. Please enjoy (in wry fashion) The Ayn Rand School for Tots.

Well, back with more on Alpha School. A lot has happened in the last 24 hours, and I am very tired (so please forgive the inevitable grammar mistakes).

The main event is this: Yesterday, Alpha School indirectly confirmed the entirety of my reporting, and promoted a microsite for its bootcamp. Let’s dive in.

Recap

Yesterday, at very same day and time that I’d informed Alpha Schools that I would be publishing my story on student life in its boot camp, Alpha School co-founder MacKenzie Price tweeted this:

To my knowledge, this appears to be the first time Price or any administrator at Alpha School has ever publicly addressed the activities of its boot camp.

Price’s tweet linked to a website with an unusual url: https://alpha-high-bootcamp.vercel.app/ It’s my understanding that Vercel is a website developer platform, and not typically something that’s user facing. Further, I asked multiple IT specialists if they could determine when this site was created, but they could not identify the specific date. Notably, however, if this site was live prior to my story being published, Alpha School never pointed me to it, despite my repeated requests for comment. Nor did Price respond to my requests on social media for clarification.

You may draw your own conclusions about the timing of Alpha School pushing this out. But please see update in footnote.1

I now turn to what this microsite contains. Later, I’ll share some student responses to my story.

I find it all so very fascinating.

Life Skills

There is a grid.

The new microsite lists 18 specific “life skills” that students must complete as “prerequisites for the deeper, more advanced work ahead” — more on this later. Importantly, each of the boot camp activities I identified in my special report are included as required skills. Each one includes a brief description of the activity and its metric of success. The header to this post highlights the Critical Thinking square, requiring that students watch a full season of Survivor and have AI grade a document of their observations.

Here are two additional examples:

Other purported life skills include Be an Expert (“Choose a topic and study it deeply enough to become the most knowledgeable person in the room”), Socialization (“Make a list of 50 people you admire and send each a thoughtful, personal message asking to learn from them”), and Storytelling & Public Speaking (“Deliver a filmed keynote…[in] one continuous take…and controlled enough that Speeko’s speech analyzer flags zero exclamation-point moments.”)

I encourage curious readers to examine the full list of life skills on the microsite.

Deep Dives (on the nightmare fuel)

The site also contains something described as “a deep dive on a few of our challenges.”

Here, Alpha School singles out three “life skills” for extended explanation: Build an Audience (the requirement that students create Twitter accounts and gain 100 followers), Receiving Critical Feedback (the biometric monitoring of student heart rates as they watch critical feedback from parents while in front of other students), and Grit & Hard Work (the requirement that students do a physically demanding “Dirty Job” – that’s Alpha School’s term).

The extended descriptions of these tasks make for remarkable albeit somewhat exhausting reading — e.g., the Building an Audience explanation is 1,833 words.2 I analyzed each deep dive using Pangram, which scored them as 61% to 73% AI-generated (for more on this, see footnote).3

Because of the length, here I selected the statements I found most interesting. This is straight quotation, with my bolding for emphasis.

Build an Audience

Our approach to social media is similar to our approach to AI. We do not believe that shielding students from a powerful technology that is essential to success in a modern world is a better choice than preparing students to use it well.

Our ultimate goal is for audience-building to serve as a feedback loop and a means for distribution.

The 100-follower goal is an objective early benchmark for whether a student is beginning to understand how to build an audience. Without an objective outcome, a student could post every day to an account followed only by classmates and then have a teacher subjectively decide that they had done a good job.4

Receiving Critical Feedback (“the heart rate challenge”)

Parents are chosen intentionally. In our receiving feedback workshop, students learn about The Mom Test by Rob Fitzpatrick, a book designed to teach founders how to elicit critical feedback from conversations even when the person they are talking to has an incentive to be supportive.

Most teenagers’ default mindset is that criticism is evidence that something is uniquely wrong with them.

If mastery depends on a teacher’s judgment, a student can always wonder whether they truly succeeded. However, an objective metric removes that uncertainty.

Grit & Hard Work (“Dirty Jobs”)

One of the mindsets we aim to instill in our students is that when something becomes difficult, students do not see themselves as victims of that difficulty.

We want students to understand that experiencing something that is difficult does not equate to being powerless, even if it initially feels that way.

The “Dirty Jobs” challenge creates a simple – and fun – environment in which students can practice this mindset.

Additional noteworthy features of the microsite

There are three additional things that I found interesting.

First, the site contains a proviso that states in full: “Both students and parents have responded very favorably to the Alpha Bootcamp tasks, but we're always conscious of the need to personalize the educational experience for students. Some students have been granted individual accommodations to modify some tasks, in consultation with their parents.”

No further explanation is provided as to when or why Alpha School will allow students to “personalize” boot camp in this fashion.

Second, Alpha School lists the daily schedule for boot camp, week by week. Here is Week 7, with apologies it’s tiny on mobile devices:

Notably, this indicates students will be spending two days “lining up” their Dirty Job, and then three days doing said Dirty Job. It also suggests they will spend all five days of Week 7 watching Survivor.

Finally, the microsite at least partially answers the purpose of boot camp and its consequences. It states that because some high school students will have already attended Alpha School elementary schools, but others will be from “very different educational environments,” it is important to bring students “to a shared starting line.” It then states, with my emphasis:

But Bootcamp is about much more than remediation…we want students to leave believing that massive acceleration is possible…We are establishing a shared threshold, rapidly catching students up on foundational skills where needed while showing every student just how quickly they can grow. When students clear that threshold, they enter the rest of Alpha High.

I will have more to say about this in the commentary.

Student Reactions

I found two student responses to my story.

First, shortly after I posted my story on Twitter, a student replied with this:

I immediately responded to ask this student who it was that made them delete their post. Unfortunately, they deleted their tweet without responding.

Second, another student re-tweeted my story and then provided extensive commentary on it. It began with this:

The student then raised a series of concerns that are too voluminous to include in their entirety, but generally questioned the overall posture of the story and contested the accuracy of my reporting. After reviewing them, and I am saying this with kindness toward the student should they read this, I did not find the factual objections to have merit.

There were two tweets however that I took particular note of:

Upon seeing these and the other tweets, I immediately responded with follow-up questions, but the student did not respond. We do follow each other on Twitter.

My commentary

I mean, you have to admit, if we train our attention solely on the curriculum of boot camp — this is pretty f’ing funny.

The students are watching Survivor to learn critical thinking? They are becoming an expert by studying something for one week and then answering 10 questions, developed with help from Google? They will develop financial literacy by learning the core frameworks from $100M Offers, whatever the fuck that is? This is what parents are paying $50,000 a year for? This is the remediation plan? This is future of education? This?

This is a joke! A very funny joke!

I am trying to convince a friend of mine, who is genuinely funny (though maybe not quite as funny as they think they are), that we should co-write a script for a television series using all of this as our source material. You’d watch Beta Camp, you would, the story of 16 teenagers who build nano-robot microducks that develop sentience and build a space rocket that shoots the confused thinkmaxxing teens straight to Mars where they land on Elon Musk and there is great rejoicing. Mountainhead meets the The Breakfast Club meets The Martian. It would be amazing.

With that said, let’s get semi-serious.

My chain of reactions after finding and reviewing the microsite: “Hot damn, Cognitive Resonance gets results,” followed by “hot damn, everything I reported has been confirmed by Alpha School publicly,” before concluding with “whew.” Whew. It was not lost on me that boot camp is only halfway over, and I have been wondering what other fresh hells might be in store for these students, and by extension me. But although the newly revealed “life skill activities” are profoundly stupid, they are not sinister. Alpha School appears to have run out of nightmare fuel. Though, I saw students tweet that there are (or were?) 21 boot camp activities, and the microsite only lists 18.

One does wonder about that gap.

Then there’s the metrics the metrics the metrics. I’ve corresponded with a few people who have some inside perspective on Alpha School and Joe Leimandt, the billionaire software-developer-turned-private equity-whatever-turned-“school principal,” and it’s become clear that for Alpha School, truth and value do not exist absent a Key Performance Indicator. The problem is that not all of life revolves around KPIs (unless they get their way!), so you end up having to concoct the absurd sorts of measures we see here, like strapping a tape recorder to a teenager as they plunge a toilet or whatever. I mean, the contempt for “teacher subjective judgment” is just dripping off the page. Or rather, the microsite on Vercel. (Still funny.)

Slightly less funny, though, are the core academic performance requirements, and in particular the SAT scores, the very first “life skill” listed. Unless they let kids use AI to take the test — and they very well might! — requiring a freshman student who scores a 1000 to jump 200 points within eight weeks...well I have many questions, because that is well-nigh impossible without cooking the books (very possible here!), especially given the lack of genuine academic rigor that permeates throughout everything that is Alpha School. Remember, according to Alpha School, students stay in boot camp until they hit all their objectives, so who knows how many of them will have to go beyond eight weeks. And then what will they do to fill the time when not cramming for their tests?

Maybe they’ll have to start watching Alone?

To the students at Alpha School, including those who now subscribe to this Substack, I am very worried about you, but I also kind of adore you. Specifically, to the student who cracked the joke but then was made to delete it (god please tell me by who), you are alright my [dude/dudette/nonbinaryette], you are behaving exactly as you should, and if I had to bet on the one Alpha student I’d most like to get an alcoholic beverage with in the future — WHEN YOU ARE OF AGE — it’d be you.

Now to all of you, please pay attention to this sage advice:

Yes. Yes! All of this! Stay alive. Break that thing. Be teenagers. Stop posting about “normal school,” be cool, be weird, be funny, be kind, make liberal use of the alleged policy that allows for exemptions from the stupid shit the Alpha School adults have cooked up at their all-staff sticky-note-bonanza “design thinking” sessions, ignore that bullshit about your default mindsets, read Ender’s Game and consider its central theme about children being deprived of their agency and their childhood (skip the movie), read James Baldwin and Audrey Lorde and poetry and fiction for god’s sake as it is where our imagination resides, don’t obsess over money, don’t obsess over status, did I mention be kind?, cultivate empathy and humility, and please recognize that this is one of the few times in your life when you really get to decide who you want to be, to shape your character, and strive to make that person extraordinary, not because of what you accomplish, but rather because of what you nurture within your soul.

And I’d still love to talk with you, so hit me up. In the meantime, stay off the vape!

Subscribe now

1

On Twitter, a user has posted what appears to be a screenshot of a comment by MacKenzie Price on Substack dated August 14, 2026 that appears to include a link to the microsite. I am trying to independently confirm its authenticity, but to my eyes it appears genuine — and I’ll admit combing through Substack comments is reaching the end of digital sleuthing skills. It would make sense for this site to have been published in mid August, as that’s when boot camp commenced. What makes no sense whatsoever is Alpha School’s PR rep describing the screenshots I sent as “mischaracterizations” and failing to point me to this site. And the fact remains that MacKenzie Price only tweeted about it on Tuesday.

UPDATE AFTERNOON 9/9/26: The Price comment has been sourced. So, it appears that Alpha Schools had the microsite prior to my story, but refused to share this information with me. And that’s really the most important part, the sharing.

UPDATE ALSO AFTERNOON 9/9/26: We live in information dystopia. Moments ago, another Twitter user found information that may indicate the site deployed a few hours before the story runs. Who knows. The important thing is the cult factory.

JFC UPDATE LATER AFTERNOON 9/9/26: Someone has sent me this screenshot and suggests it indicates this page on Alpha School’s main site was created on September 7. It links to the microsite. So there.

I have been writing essays here for just over two years, and what follows is the strangest I’ve composed to date—by far. It’s a report on life in the first weeks of high school at Alpha School, based on what students are sharing on social media. Because what they describe is so profoundly weird and concerning, I am dividing this into four sections:

1. Reporting on Alpha School’s boot camp

2. Alpha School’s response

3. Reactions to Alpha School’s practices as described in this story

4. My commentary

Buckle up, grab a beverage, this is a long journey.

Before we get into things, there’s some important context I must share. First, if you’ve read or know anything about my work, you know I do not subscribe to the vision of Alpha School’s leaders regarding the role of AI in education. What’s more, because Alpha School is a very small private school network that serves families that can afford ~$50,000 in annual tuition, until the last week I’ve paid very little attention to the specifics of its model, for the simple reason that the world they inhabit looks nothing like my own, nor that of most families in the US (or worldwide). The press may be captivated by their story, but I am not. Indeed, after two decades of working in education and seeing prior incarnations of Alpha School-like efforts fail spectacularly, I’ve focused on other matters.

All this changed midweek last week when my phone blew up with people sending me messages related to Alpha School’s official Twitter account, and tweets that appeared to be written by its students. At first, both my friends and I thought it might all be part of an elaborate prank, because the style and substance of what was being shared was so bizarre, it seemed impossible that they could be genuine. Further, many of the tweets appear to be generated in whole or in part by AI, adding another layer of difficulty in separating fact from fiction.

However, after conducting an independent investigation, I believe that the accounts are legitimate, and that taken as a whole, the tweets convey aspects of student life in the first weeks of Alpha School’s high school, the period called “boot camp.” Although tweets are obviously public, and Alpha School itself is promoting them through its Twitter account, in an abundance of caution I have blurred out the student avatars and names here.1

Please note further, while I have many critiques aimed at Alpha School’s leadership and the vision of education they are promoting, the students are blameless, they are simply teenagers trying to make their way in the world. And as is so often the case in education, I am grateful for their honesty and insight, and for sharing details of their experience.

Let’s turn to that now.

Reporting on Alpha School’s boot camp

1. Alpha School’s official Twitter account created and promoted a list of student accounts on August 24, 2026.

You can access the list here, and as of September 7, 2026, there are 46 affiliated accounts. Although Alpha School refused to confirm or deny that @Alphaschool is the organization’s official account, the gold check next to the handle indicates that “this account is verified because it’s an official organization on X.”

2. Alpha School’s students are apparently participating in a “boot camp” for their first eight weeks of high school. As part of boot camp, it appears they are required to post on Twitter and obtain at least 100 followers.

Numerous students report that they are currently experiencing what Alpha School calls “boot camp,” which is not something I found described anywhere on Alpha’s official website, or anywhere else. The students share that, as part of boot camp, they are required to join Twitter and obtain at least 100 “authentic” followers by early October 2026. Most of the student accounts appear to have been created in August this year, when boot camp commenced.

The tweets below are illustrative. Note the last student stating that “building an audience is what we’ve been working on for the past seven days at Alpha high.”

3. As part of boot camp, Alpha School students apparently have their heart rates monitored as they watch recorded videos of their parents critiquing them. This activity takes place in front of their classmates, and students must keep their heart rates within 10 percent of their baseline to pass.

Numerous students refer to a boot camp activity that, as the heading describes, involves biometric monitoring of students as they watch pre-recorded videos of their parents providing critical feedback. This is done in group settings with student peers. Below you will find multiple tweets of students describing this activity (and there are more like these).

Please note the first student describing their hands shaking “the entire rest of the afternoon” after doing this.

4. It is unclear what consequences Alpha School students face if they fail boot camp.

There are multiple posts by students suggesting they found out about boot camp as the school year commenced. At least one student indicated that they must complete boot camp activities in order to stay enrolled in Alpha School:

In contrast, one student said that their enrollment was not contingent upon completing boot camp. Instead, they described that certain privileges are denied to students until 21 separate “life skill” activities are completed. Specifically, students are (1) denied access to their phones during the school day, and (2) denied permission to go off-campus during breaks. “There really are no other consequences,” this student said.

A PR representative for Alpha School refused to answer any of my questions regarding boot camp or its consequences. I did speak with someone with extensive career experience working in the field of independent schools (what most people call private schools), and this person confirmed it would be highly unusual for Alpha School to dis-enroll students after eight weeks — in their words “typically, once [an independent private school] enrolls a student, they’re in, barring unforeseen circumstances.”

Dan Meyer has reported in more detail on Alpha School’s admission policies for its independent private schools.

5. Another Alpha School boot-camp activity apparently requires students to do unpaid labor cleaning restaurant kitchens or gas-station bathrooms. Students audio record themselves as they do this work, and if they complain, they fail.

One of the apparent life skills that students must cultivate in boot camp is “grit and hard work” — as noted above, this can apparently include cleaning toilets, scrubbing dirty restaurant kitchens, or mucking horse stalls.

Two students describe this activity as follows:

6. At least one student at Alpha School has expressed concern about boot-camp activities to MacKenzie Price, Alpha School’s co-founder.

On July 14, 2026, a few weeks before Alpha School’s boot camp commenced, Alpha School co-founder MacKenzie Price tweeted this:

She received several responses, all seemingly from adults, and all roughly within 24 hours of posting. However, last week on the night of September 3, I discovered a single tweet from a student at Alpha School that appeared to have been posted within the previous day:

That same night, I responded to this message to ask this student to elaborate on their concerns, but the next morning, the tweet had been deleted.

7. During boot camp, Alpha School students apparently engage in sustained discovery learning. This is described by some students as a “core life skill.”

Numerous students report that, as part of boot camp, they are required to complete one unit of an Advanced Placement course and receive a score of 5 on the end-of-unit exam (my spot checking suggests a full AP course usually contains ~10 units). Students must accomplish this within five days, and they must do so entirely through their own self-directed learning using AI and information on the Internet, what in the education field is called “discovery learning.”

These tweets are illustrative:

The emphasis on “learning to learn” as a “core life skill” in boot camp is particularly striking given that Alpha School has engaged Dr. Carl Hendrick as consultant to its activities (described briefly here). Hendrick is longtime vocal critic of discovery learning — as one example, in 2020 he co-authored How Learning Happens, a book that includes chapters titled, “Why independent learning is not a good way to become an independent learner,” and “Why discovery learning is a bad way to discover things.”

In a more recent interview, Hendrick said this about Alpha School’s future AI plans, with my emphasis:

I contacted Hendrick and requested public comment from him regarding Alpha School’s widespread use of discovery learning, as well as its use of biometric monitoring. He declined to comment. He did state on Twitter that, “adult educators with huge social media followings publicly shaming kids to criticise a school they don’t like is a new low...But if you’re the good guys saving the world from the ‘big-tech bro’ bad guys then anything is permitted.”

8. Alpha School students often compare their experiences and academic achievement to what they call “normal school.”

Here are some quotes from students attending Alpha School about their experience in comparison to other schools and students:

• Your child isn’t learning anything at school, because their lessons are paced for the slowest kid in the room.

• While kids at normal schools take a bus to a museum, we’re getting split into 10 tribes to build tarp shelters with no tents.

• My school may give 50k to a group of students that win our take on survivor. At

  @Alphaschool we value learning what “normal school” won’t teach you.

• I learned how to belay another person that was really cool. Over all thankful towards the @Alphaschool team for putting this together for us, and giving us the opportunity that we wouldn’t get in normal school.

• Most students sit in class for 7 hours and forget half of it by Friday. At Alpha School we finish our core academics in about 2 hours on a learning platform and spend the rest of the day on real skills like building audiences, giving feedback, and working in teams.

• I use AI in @Alphaschool every day and it is proven to 2x our learning speed compared to normal students. Why do protesters, adults, and the mayor of NYC all oppose something that will revolutionize education across the World?

• My team of high schoolers just scored higher than most Harvard MBAs at their own leadership simulation.

• Today at @Alphaschool, my team proved that we are smarter than 85% of Harvard students by climbing Mount Everest with 4 other students.

• I just got 98th Percentile Nationally on my SAT...As A Freshman. It's been proven that most parents lock their kids in rooms to improve their SAT scores; that doesn't work.

• @Alphaschool’s incoming freshman score better than the average senior. Today I took my first ever practice SAT today and I’m already in the high performers.

• Here at @Alphaschool we are learning two times as fast as other teens. We need to hit mastery, which is 90% or over.

• Imagine going to a school without teachers, homework, and grades. That's my reality everyday.

• I just self taught myself AP World history without any teachers.

9. Many boot-camp activities at Alpha School appear focused on business education.

Many Alpha School boot camp activities feel akin to what graduate students might experience pursuing an MBA.

Among other things, Alpha School students participate in a three-day “real world Survivor” wilderness activity. The curriculum includes Radical Candor, Atomic Habits, Legacy, Extreme Ownership: How U.S. Navy SEALs Lead and Win, Grit, Courage to be Disliked, The Culture Code, and the essay 1,000 True Fans. Students also participate in the Harvard Everest simulation, an activity sold by Harvard Business School. Many of the students describe starting their own companies, such as a basketball shoe manufacturer or a “car-flipping” business, or doing something in tech, such as creating a “prompt optimizer.”

In reviewing more than 40 students’ accounts and hundreds (perhaps thousands) of tweets, I did not see a single reference to a work of fiction.

Alpha School’s official response to this story

Although my essays typically are more op-ed than straight news reporting, in this case I felt it prudent to offer Alpha School the opportunity to respond to, and if necessary clarify, whether the activities described by its students are accurate representations of its boot camp. Alpha School’s website does not list a direct press contact, but I was able to reach Daphne Ortiz with Statement PR, which represents Alpha School.

In my initial email, I asked for information on whether Alpha School was requiring students to use Twitter, the structure of boot camp and consequences for failing, and the biometric monitoring of students’ responses to parental feedback in group settings. I included some of the screenshots listed above. Ortiz responded with the following statement, reprinted here in full:

I replied as follows:

Thanks for the prompt response Daphne. Just to be clear, these tweets were all found via the Twitter account purporting to belong to Alpha School, and as noted, many of the tweets say that they are tweeting as part of their coursework, and further, many state they are supposed to gain 100 to 1000 followers by October 1.

So I guess my follow up question is, did Alpha School weigh the ethical issues you are suggesting I or any other journalist should follow, and nonetheless decide to promote these student accounts through its official Twitter handle?

Ortiz did not respond further.

Neither MacKenzie Price nor Joe Liemandt (another co-founder of Alpha School) responded to my requests for comment via social media.

BREAKING (September 8, 2026)

Alpha School appears to have launched a new microsite explaining its bootcamp, which you can find here. I strongly suspect, but cannot yet confirm, this site was launched after I informed them of my forthcoming story (I requested comment on September 3.)

Based on initial review, the site confirms every activity reported on here, and then some. I will have more to say about what it reveals in the days ahead.

Reactions to Alpha School’s practices as described in this story

Larry Ferlazzo (Semi-retired veteran teacher and Education Week columnist):

Requiring students to post publicly on Twitter is inexcusable. Do teachers really want to create more opportunities for students to be bullied or receive unfair criticism than they already have on their own? It’s likely a public school teacher who assigned students to do something like this would be immediately suspended.

I believe it’s important for teachers to create the conditions in their classroom so that students can feel powerful, and feel positive and proud about themselves, their families and their communities. It’s concerning to me that a fair number of the comments purportedly coming from Alpha students show them bragging that they are BETTER than others. That attitude seems to be antithetical to what most educators want to promote.

I am floored by the irony of Alpha School promoting “discovery learning,” if that information is accurate. For many years, I have responded to critics of discovery learning by stating that I knew of no educators actually doing it, and that it was a straw man used to support the idea of direct instruction being the one true form of teaching. To see some of those critics supporting Alpha School, which appears to be the first school I’ve seen actually doing discovery learning, is ironic indeed.

John Warner (Writer and educator):

We're at an inflection point where there is a rising fever and something must break it, either we will blast through and decide school is a kind of weird entrepreneurship boot camp...or something for human beings.

Adam Boxer (Science teacher and director of education for Carousel Learning)

It appears that students are being asked to make Twitter accounts as part of their school curriculum and work. This strikes me as very odd indeed. Social media is not a safe space for children, and encouraging them to build a following increases their online vulnerability to predatory actors.

If a teacher in an English school asked their students to create social media accounts and seek a larger following, it would very likely result in an immediate safeguarding referral directly to the principal/headteacher of the school, followed by escalations to the LADO (Local Authority Designated Officer), and potentially the Teacher Regulation Agency. It is unlikely that any teacher or educator who supported a practice like this would be employable at a school in England.

Dylan Kane (Veteran teacher and writer)

The real talent of folks at Alpha School is marketing. More specifically, they are good at marketing to a type of highly-online, Silicon-Valley-adjacent, anti-traditional-education world where they've made their niche. Their marketing is full of high-level claims about what is possible in education and very few details about exactly why what they do works. They're impressively apathetic about the details of their academic program.

I'm a teacher, and I will tell you that most days teaching feels really hard. I would love to get better at it, or to find a better way to structure schools. But Alpha School seems totally uninterested in helping the broader education system. Their disdain for people like me is obvious. Their priority is saying stuff that sounds clever on the internet.

Audrey Watters (Educator and writer)

“You don’t need a weatherman to know which way the wind blows” — Bob Dylan

AlphaSchool’s promises have always been mostly snake-oil: this idea that artificial intelligence can deliver high-speed, high-quality education is mostly the stuff of science fiction, not reality.

But now, for AlphaSchool to embrace social media and “AI” as the winds shift, as courts weigh whether the companies that build these products are liable for actively harming children, seems not just out-of-touch with public sentiment but actively dangerous and damaging to the students in its charge. Indeed, encouraging AlphaSchool students to use Twitter isn’t simply asking them to give up their privacy or to offer themselves to software systems for algorithmic evaluation and manipulation. Elon Musk has openly touted his sites as crucial for right-wing indoctrination; and more damningly, Grok and X have been accused of creating and spreading CSAM [child sex abuse material].

What we witness with AlphaSchool isn’t a matter of simply ignoring the weatherman, fingers-crossed hoping that the sun will continue to shine on the chosen few and that the winds will shift again in industry’s favor. Rather, it’s pretty clear that there’s an absence of any moral compass at all.

Lila Byock (Parent and founder of Schools Beyond Screens)

At a moment when major public school districts like NY and LA are leading by example to protect students from predatory tech companies, it’s disheartening to see for-profit Alpha School clinging to such an outdated vision of “future readiness.” Education should empower children to achieve their full human potential, not count followers on a dying social media platform. Kids’ dreams are too big to fit into 280 characters.

Jane Rosenzweig (Writer and educator)

I reviewed some of the supposed “student” posts, and as a writing instructor, I found one particularly sad: The “student” wrote about using a“web optimizer” to generate a prompt to ask ChatGPT for essay ideas for his friend. This yielded “five solid thesis options” and a finished essay “before lunch.” It’s hard to see where the learning is in this approach, and the message the “students” seem to be getting is that writing is something to be produced as quickly as possible, rather than a way to figure out what you think about something.

Anne Lutz Fernandez (Writer and educator)

Most of what I have to say about this is unpublishable. If the teenagers at this school are being subject to much of what has been posted on Twitter via the Alpha Schools account, someone there needs to be reminded that teenagers are children. They are not soldiers in an army, they are not adults at a business school, they are not workers in a pressure cooker environment.

I spent 10 years on Wall Street in mergers and acquisitions, and my 20- and 30-something colleagues and I were not treated as these students appear to be. Especially dark is the biometric monitoring, which sounds like something a CIA trainee would be subjected to. Students in “normal” schools face serious pressure to achieve and are increasingly over-surveilled. This takes those elements to a new and unheard of level – and especially if the leaders of Alpha School aim to influence how schooling is done across the United States, I’m profoundly disturbed.

Shriram Krishnamurthi (Professor of Computer Science)

If this situation accurately represents what is happening at the school, I feel most for the students who have been forced into an unforgivingly brutal public sphere. It's one that has undoubtedly been useful as a source of knowledge but is also a shiv that periodically sticks in you. Doing that for what is clearly its real purpose — to raise the school's visibility — feels like a violation of sacred educator norms.

Stephen Fitzpatrick (Veteran teacher and writer)

My interest in Alpha is mostly as a counter-point to our current system of education — they are exploiting a system that runs on tests as metrics of success. There are absolutely kids who can get a 5 on an APUSH [single unit] exam in five days — does that mean they’ve gotten the same experience as a kid who had his eyes opened to the history of this country in a class on American history over nine months but didn’t get a 5? Of course not! But the state rewards the 5’s.

If Alpha wants to push its test scores as their main measure of success, how they arrive at, achieve, and report those definitely need to withstand scrutiny.

Andy Liddell (Attorney with Ed Tech Law Center)

If the reporting accurately describes students’ experiences at Alpha School, I am profoundly concerned for their wellbeing.

In my work with the EdTech Law Center, I represent children who are harmed by the technology they are made to use at school. If you can imagine the worst things that can happen to a kid on the internet, they’re happening every day to kids on the internet at school.

But rarely are schools actively putting kids in harm’s way as enthusiastically as Alpha School reportedly does: not only are students being flung into commercialized online spaces that are havens for scammers and predators and powered by hostile algorithms, they are being surveilled, quantified, and publicly humiliated by their own school leaders in ways that seriously risk their mental health.

Alpha School is not the future of public education. I question whether its unproven model even has a place in present-day private education.

Callie Lowenstein (Educator and founder of Acuarela Biliteracy Works)

The practices described here are an almost perfect distillation of dehumanizing pedagogies, a reminder of why the Humanities have their name...and why they are so important. I just want to sneak a little Mary Oliver poem into the pocket of every teen in this program, a little Audre Lorde here, a little James Baldwin there. I wish for them:

To learn something by being nothing

A little while…

My commentary

I don’t even know where to begin.

When I shared screenshots with my friends in education about Alpha School biometrically monitoring their students, their first reaction was typically “WHAT?” or “WTF,” soon followed by “this can’t be real” or “this is The Onion, right?” In fact, one friend of mine, a close observer of Alpha School, finds the whole situation profoundly funny, and with good reason. If I were to pitch HBO on a new series with the logline “Silicon Valley meets My So-Called Life,” the scenes would be taken straight from Alpha School’s boot camp. Everything about this feels like sublime parody. (This might be true about life in general nowadays.)

But what I cannot fully convey is how profoundly surreal it was to spend several days immersed in online conversations about boot camp, and for the first time contemplating what is actually happening at Alpha School right now. Because it seems to me that one of only two possibilities must be true:

(1) The accounts and tweets are not authentic. It’s possible that Alpha School has created de facto “Potemkin students,” meaning that, while the accounts may be affiliated with actual students, the tweets are largely being directed or steered by Alpha School employees. A PR campaign using student avatars, essentially. That would at least explain why so many of the messages feel so radically inauthentic to everything I know about teenagers (and here I must reiterate that I’ve spent two decades working in education, so I’ve met my fair share). When I see a purported 14 year old stating, “your child isn’t learning anything at school, because their lessons are paced for the slowest kid in the room,” I am beyond flummoxed, because kids don’t usually talk like this — who is the parent audience they are imagining? Why do they care? Why are they spending their time doing this?

So while it would be abhorrent to confirm that Alpha School officials are using student accounts for propaganda purposes, I’d find it relieving, too. Adults behaving badly in education? Not my first rodeo. However, I don’t think that’s what’s happening.

(2) The alternative is that these accounts are authentic, and the tweets accurately reflect the activities of Alpha School’s boot camp, and student dispositions more broadly. As noted in the introduction, many of the tweets are obviously AI generated (some even are labeled as such by Grok), but you still have to prompt a chatbot to generate its output (for now anyway). Plus, the student who messaged MacKenzie Price to raise her concerns — can’t explain that one as a bot. So as I’ve said, after investigating independently, I personally believe the accounts are legitimate, and real students are posting to them. But if that’s the case…

…we have to take semi-seriously the question of whether Alpha School is a cult? A set of indoctrination camps for the airport-book class? What the hell is happening here?

This is A Clockwork Orange meets B.F. Skinner nightmare fuel, what with Alpha School student heart rates being biometrically evaluated and these students expressing triumphant superiority over the out-group “normal school” children and these students engaging in ritualistic and some might say degrading bonding activities and these students repeatedly praising the virtues of their one special school, over and over and over. “At my school we do this, at Alpha we do that, my school Alpha School my school Alpha School my school Alpha School,” the chant of the Hare Krishna as voiced by Veruca Salt. Daddy I want to go to an Alpha School nowwwwww.

And then on top of this we have Alpha School cloaking itself under the banner of “the science of learning,” a term I helped popularize, while promoting radically extreme behaviorist practices such as heart-rate monitoring (!) and pure unadulterated discovery learning by combing the Internet (!!!) and an affiliated “learning scientist” is promising us that in the future AI will be used to monitor student behavior, including “latency” of physical reactions (!!!!!!!) and oh by the way the co-founder is apparently aiming to reach one billion students…well, I’m rarely at a loss for words, but what can be said?

This is madness. Actually, my young daughter said it best after hearing these boot-camp activities described to her: That’s not okay.

No it’s not. None of this is okay.

In the course of writing this essay, a few people expressed nervousness about my reporting on students using social media. I get it, and I agree students should not be forced to do this (and please see the footnote), but it’s precisely because these are real kids living in the real world that I’ve written this story.2 The most agonizing post I read is the one shared in #6 above, the student who messaged MacKenzie Price to express their concern that the activities being undertaken in boot camp were “faulty” and involved skills or morals — please linger on that word choice — that clash with other, deeper values. Price is fond of hand-waving away Alpha School critics that she sneeringly labels “haters,” but this was one of her own students speaking up, clearly not a hater, just a teenager grappling with a very confusing boot-camp experience and sensing, rightfully, that something profoundly fucked up was happening.

In reviewing drafts of this story, a few folks also suggested that it wouldn’t be “the worst thing in the world” for teenagers from privileged families to have to scrub toilets or clean horse shit. Here’s the thing, though: (1) Numerous students expressed complete surprise by this activity — and indeed, boot camp itself — and as far as I can tell, nowhere does Alpha School publicly advertise such activities are part of its program. I find that telling. (2) There is extensive academic research on how selective groups — such as cults, gangs, fraternities, and emerging religious sects — make use of “shame and stigma” rituals that include deliberately demeaning initiations so as to enhance commitment to the in-group and its norms. In my view, these boot camp activities are intended as bonding rituals. As such I would be quite surprised to discover that Alpha School continues to make students clean toilets or horse stalls beyond boot camp, but if its leadership wants to confirm on the record that they do, I’m all ears. Though then we might have to have a different conversation about child-labor laws.

Will any of this derail the Alpha School Hype Train? Well dear readers, this is where you come in. I was grateful to see my recent The 50 Percent Problem essay “do numbers” as it became the second-most popular post I’ve published — my thanks to all of you who read it, commented, and most importantly shared it. Now, I’m asking you to do the same with this one. Let the word go out:

What’s happening at Alpha School should shock the conscience, and going forward, everyone who covers them should ask hard questions about the student experience at Alpha School as described here. Journalists: Please investigate, interview, dig into the numbers, review the failed charter applications (and there are many), do whatever is necessary to bring to light what’s really going on.3 And for everyone else, myself included, it’s time to train our critical gaze upon the world that Alpha School is “optimizing” for, the culture it aims to foster, and the values it hopes to inculcate in children. Education is not thinkmaxxing.4

In a world where attention is precious, I vowed to give as little of mine as possible to this grotesque little enterprise of the elite. But I’ve 2xed my own learning about what they are doing.

Alpha School, you have my attention now.

My thanks to the many people I consulted with in reporting this story — as always, all errors and mistakes are mine alone.

Subscribe now

In 2024, Laurence Holt of the XQ Institute published an essay titled “The 5 Percent Problem,” a term that quickly became common parlance within the ed-tech zeitgeist. Holt argued that when it comes to measuring the student-learning outcomes of various ed-tech programs, the results often only apply to the five percent of students who “used the program as intended. The other 95 percent see minimal gains, if any.”

Holt’s primary focus was on online math programs, and the central concern running throughout his essay involves whether and how students might “get the recommended dosage” of using various ed-tech tools. Whether the problem arises from unmotivated students or unenthused teachers or disparate access to tech or some combination thereof, the challenge when framed this way is getting students to opt in to using a particular technology.

Generative AI does not suffer from this problem. Uniquely perhaps in the history of ed-tech, AI has been broadly embraced by students worldwide, so much so that the phrase “AI is inevitable” has become commonplace in the education discourse. There are pockets of resistance, of course, and over the last six months we’ve seen mounting opposition to AI across multiple vectors, most prominently with data centers. Nonetheless, we all know that students are using AI frequently. “Dosage” is not an issue.

In the nearly four years of time that’s passed since ChatGPT was commercially deployed, however, we have suffered from a lack of high-quality empirical research on the impact of students using generative AI. There are notable exceptions—such as these—but a recent research landscape analysis out of Stanford indicated that, out of more than 800 studies of AI in education, a mere 20 employed true causal measures. Meanwhile, at least one prominent meta-analysis purporting to show massive learning gains stemming from AI has been retracted, though not before being viewed 400,000 times. The field is a mess.

Perhaps this is because studying AI in “the real world” poses serious challenges. We know that AI tools are free and widely available to students. We also know that many are using them after school hours for various purposes, including to help with their schoolwork. But quantifying this is very difficult, because it’s hard to peer into the home lives of children. And it’s equally if not more challenging to connect out-of-school behavior to measurable learning outcomes, the data sets are not easy to link. Accordingly, research on AI’s education impact has tended toward lab studies or small-scale evaluations.

No longer.

Last week, a study titled “The Generative AI Learning Penalty: Evidence from Chinese Secondary Education,” authored by David Strömberg, Victor Lei, and Yanhui Wu, went viral after The Economist published data visualizations from the research paper. The preprint came out in June—not sure how I missed it—and uses data from approximately 27,000 Chinese students in grades seven to 12 to explore a straightforward question that has preoccupied me for several years: “How does self-directed use of generative AI affect cumulative learning over time in ordinary school settings?”

We will get into the details shortly, but here’s the headline summary of the results:

• Between 2023 and 2025, approximately 80 percent of all students started using generative AI, and after just five months of use, approximately 50 percent of these students engaged in “full homework outsourcing,” meaning, they essentially stopped doing their homework independently and used AI instead.

• Over time, this caused significant learning loss as measured both on monthly closed-book exams and via the comprehensive high-stakes exams that China uses to determine high school and college placement for students. On those tests, by the end of June 2025, overall student performance fell by 24% on the former and 18% on the latter, a massive decline.

• What’s more, there is no evidence—none—of any corresponding learning benefit arising from students using AI.

• As such, the researchers state plainly that “our findings show that generative AI, which is likely to become a prevalent technology for education, has a substantial negative impact on student learning.” (My emphasis)

Put simply, we now have rigorous, empirical evidence on the real world, long-term impact of students using AI. This research indicates that AI substantially harms learning for half of all students, what the researchers call the “Generative AI penalty,” and provides no discernible benefit to the other half. Again, by the time this study concluded, 50 percent of students using AI with regularity had fully outsourced their homework to generative AI. For these students, when at home, they just stopped thinking about their schoolwork.

I hereby dub this the 50 Percent Problem. Unlike the Five Percent Problem, the 50 Percent Problem is a measure of harm stemming from widespread adoption of AI by students, rather than their failure to use this technology. It is my conservative estimate of how many students are being directly harmed by using these tools.1 As I’ve said repeatedly for several years, generative AI is a tool of cognitive automation. It’s both predictable and tragic that students are learning less because of it.

This new empirical research provides us with a clear measure of the degree of AI’s educational harm.

Let’s turn to the data itself. In what follows, we’ll be looking at the performance of three different groups of students across three different learning activities.

As to the students, we first have the baseline group of student results from the period prior to the introduction of generative AI—these students are labelled “Pre AI,” and charted in green. Next, we have the admirable group of students, all 19 percent of them, who did not use generative AI at any point between 2023 and June 2025 when the study concluded—these are labelled “Never AI,” and charted in blue. Finally, we have the remaining 81 percent of students who adopted AI and used it for their schoolwork—these are labelled “Used AI,” and charted in red. Of note, these students did not all start using AI at the same time, adoption phased in over time (which allowed the researchers to uncover some interesting things that we won’t go into here).2

As to the learning activities, we’ll start by looking at the amount of time that students spent on their homework; the researchers were able to track this because students had to log in online to download their homework and upload it upon completion. Then, we’ll look at how students scored on this same homework. Last but definitely not least, we’ll examine how student performance changed on monthly closed-book exams, as well as China’s very high-stakes high school and college entrance exams.

Both the underlying research paper and The Economist present a series of helpful data visualizations of this, but without the underlying raw numbers. So I contacted the researchers who conducted the study and David Strömberg (the lead author) graciously agreed to provide me with the underlying histogram data I used to create the three animated GIFs you’re about to see.

With that as backdrop, we can explore three specific questions.

1. What is the impact of AI on the amount of time that students spend on their homework?

The results here are unsurprising—AI significantly reduces the time students spend doing homework. On average, both the Pre AI and Never AI groups spent about an hour doing it, across a range of 50 to 80 minutes. In contrast, within the Used AI group, most students spent around 45 minutes, and some even cruised through in 25, presumably the minimum amount of time it takes to cut-and-paste answers out of a chatbot.

This is AI as tool of cognitive automation working exactly as intended.

2. What is the impact of AI on how students score on their homework?

Here again we see that Pre AI students and Never AI students have near-identical results, as we’d expect. But not so with the Used AI students—now, we see a huge shift to the right in purportedly “positive” outcomes, with scores far higher than even the most studious Never AI student managed to achieve. (Note the scores here were normalized to make the average score a 100 (not the maximum), so a score of 130 means “30 percent above the average,” essentially.)

To restate, the Used AI students studied less yet scored better than their peers. That’s a pretty sweet deal for them, but it’s worth reflecting on the broader implications of this within schools. When I talk to students about AI, one thing I hear them say frequently is that they don’t want to be played for chumps (my term, not theirs). Meaning, if they know their classmates are using AI and getting better grades as a result, even those inclined to resist AI may feel trapped into using it, just to keep pace.

It’s a cognitive race to the bottom.

I’ll also add that homework scores are obviously a very imperfect measure of student learning. In my view, a great deal of education research, and certainly the studies often promoted by ed-tech vendors, use comparable “point in time” data of supposed learning akin to what we see here. This is understandable, to a degree—it’s very difficult to information on long-term learning outcomes. But what ultimately matters in education is building durable student knowledge. The seductive danger of AI exposed here is that, by using AI, students may falsely have believed everything was proceeding swimmingly. In this sense, AI fosters a mental masquerade, scores go up as actual learning goes down.

Now to remove the mask.

3. What is the impact of AI on how students perform on closed-book exams?

Here is where the proverbial rubber meets the road.

Consistent with the two previous data sets, we again see near-perfect alignment between the Pre AI and Never AI students on the monthly closed-book exams administered to students. But now the adverse impact of AI is laid bare—just look at that shift to the left. The Used AI students are scoring at levels far lower than their Never AI peers, indeed, many score far lower than anything recorded prior to AI existing.

I’ve labeled this the Deadweight Learning Loss to underscore the volume of harm caused by the generative AI learning penalty. What’s we’re seeing here is unambiguous evidence of a decline in overall student performance that grew over time as students adopted AI. In fact, the average decline was 20 percent with a 1.4 standard deviation (SD). Although it’s far from an apples-to-apples comparison, this vastly exceeds the estimated learning loss in the US after the pandemic (approximately .25 SD in math and .13 SD in reading).

What’s more, over time this Deadweight Learning Loss had significant adverse consequences for students on the comprehensive high-stakes high school and college entrance exams that China administers to determine student placement.3 For students who adopted generative AI two or more years prior to being tested, the estimated negative effects are 24 percent (1.5 SD) for the high school exam, and 18 percent (1.3 SD) for the college exam.

If you are a parent with a child in junior high or high school who uses generative AI, this data should terrify you. This isn’t about academic integrity, it’s about a generation of kids being told AI is “inevitable” and “the future” and acting accordingly, in a societies where we’ve yet to develop firm norms about what’s acceptable to do with these tools within education. The upshot is that students are using AI in ways that are harming their cognitive development and foreclosing their life opportunities. Harms that are not easily remediated, if at all.

This is the 50 percent problem. Generative AI is cognitive cancer and we are doing next to nothing to stop it from spreading.

I’ll now address a few objections.

First, I want to credit The Economist for putting this research on my radar, and for raising broader attention to its harrowing findings. Yet, remarkably, in the brief article accompanying the research data we’ve just covered, the anonymous magazine author suggests that “AI can boost learning productivity but only for those who use the technology intelligently.”4 In similar fashion, Blake Richards, a researcher at Google who works “on the intersection of machine learning and neuroscience,” argued on social media that AI is not really the problem here, because “if you control for how long students spend studying, then students using AI actually perform equal or better” than those that did not.

Motivated reasoning is a powerful force. Here, of course, “controlling” for how long students spend studying erases the major findings of this research. But let’s leave that aside, and probe whether this argument can be justified on its own terms—does this study suggest that AI can boost learning if used “intelligently”?

As best I can tell, this claim is premised on this graph that correlates homework completion time with exam results:

If we ignore all that pesky data on the left side of this chart (which of course we shouldn’t), and just focus on the overlap between the Generative AI students who continued to study for the same duration as the No Generative AI students, it’s true there’s no major gap between them. Of course, note that there’s no discernible benefit to using AI either—apart that is from the spike at the very top end.

So what’s the deal with the spike? Might we cling to it as proof of the “promise” of AI? Well, here’s a good lesson on why one should always be careful eyeballing charts without access to the relevant underlying data, because when I asked Strömberg how many Used AI students persisted in studying for at least 75 minutes, he told me this (via email, with my emphasis):

It is very rare for students who have adopted AI to spend 75 minutes on homework. This occurs for only 20 students, and for each of them only in a single month (0.2% of AI student-month observations). More than five months after AI adoption, we never observe students spending 75 minutes on homework. In fact, in this group, only four students spend more than 65 minutes on homework.

Let’s get real. Given this study involved almost 27,000 students, indexing on this vanishingly small number of studious AI-using students isn’t just putting lipstick on a pig, it’s smearing its body in Revlon from snout to tail. We need to stop pretending there is some massive benefit to learning if only we trained kids properly on how to use AI. It’s fundamentally harmful.

A different and more sophisticated counterargument might proceed along the following lines: This study tracked student usage of AI stemming from its earliest days, when the tools were less capable than they are today. Perhaps relatedly, the researchers here found evidence that it was the early AI adopters most harmed by generative AI, insofar as “the estimated AI learning penalty fell from around 25 percent in early 2023 to around 16 percent by June 2025.” As such, perhaps this study constitutes the “high-water mark” of AI-induced harm, and perhaps the AI learning penalty will continue to drop. Or so we might hope, anyway.

But as the cliche goes, hope is not a strategy, and there are some problems with this counterclaim. For one thing, the rapid evolution of AI cuts both ways—we now have AI companies explicitly marketing “AI agents” to students to complete their tasks for them. Agents are even worse than chatbots from a cognitive development standpoint—at least the latter require an interaction of some sort to produce output. For another, even if the magnitude of the learning penalty continues to diminish, there’s a question of volume, too. Recall that 20 percent of students managed to resist using AI prior to June 2025. Do you think that number has gone up or down since then? I know my bet.

Finally, I can imagine someone saying I’m placing too much weight on this research—it’s just one study from China, after all. On that front, and as a self-sanity check, I asked three PhD education researchers to review the methodology employed, and all three came back with positive reviews (“it’s very good work,” said one). And China is surely the one country in the world that can match the US for AI adoption and enthusiasm, though it may surprise you to learn that AI companies in China disable their products completely during the high-stakes testing periods. (OpenAI, in revealing contrast, heavily promotes ChatGPT on college campuses during finals week in the US.)

Perhaps more importantly, if you are an AI-in-Education Enthusiast, I feel confident in saying that you will not be able to produce research of comparable rigor that shows positive long-term education impact of generative AI in real-world conditions comparable to those here. If there were such evidence, my inbox would be filled with people jamming it down my throat, trust me. And no, this single study from 2024 involving roughly 150 physics students at Harvard is not going to cut it, sorry.

The 50 Percent Problem will not disappear of its own accord. The Edu-Cognoscenti continues to chatter about learning loss related to school closures during the pandemic. Well, the learning loss stemming from generative AI appears more substantial, it’s happening right now, and it may endure for far longer. So to all the policymakers and philanthropists and so-called thought leaders who purport to take education evidence seriously, I ask you: What are you doing to prevent ongoing educational harms of generative AI? Are you doing anything at all?

We are in cognitive crisis.

I’ll close with this. As I was drafting this essay, my friend Dan Willingham published his perspective about when students should use AI. Please read it. Although he’s a tad more measured (or perhaps just realistic) about its role in education, his core contention echoes what I’ve been arguing for several years, and stems from a basic understanding of human cognition:

A tool that should only be used once you already know something is not a learning tool. We must stop gesturing at AI’s imagined potential, and focus our efforts instead on mitigating the 50 Percent Problem it’s created.

How, you might reasonably ask? It won’t be easy. But I have some emerging ideas, stemming from recent investigations into communities of technological refusal. It’s been quite the learning journey. More soon.

My thanks to David Strömberg for providing his underlying data and to my anonymous academic friends who reviewed the study and this essay—all errors mine and mine alone, of course.

Subscribe now

A friend said to me recently, “my LinkedIn is covered with references to the science of learning, you must feel so proud!”. She was thus surprised to learn that, much like my friend and collaborator Dylan Kane, lately I fall somewhere between ambivalence to concern regarding the increasing use of “SoL” in the education discourse.

Is this just my natural contrarianism? Possibly, possibly. But for reasons we’ll get to, as much as anyone I bear responsibility for popularizing the term “the science of learning,” and nowadays I see it being invoked in ways that bear little to no resemblance of what I intended. It’s become a buzzword. And while I find a small measure of satisfaction that many feel the need to nod in the direction of using scientific insight to inform education practice and policy, we’ve reached the point where SoL can mean just about anything, which is to say, it is verging on meaning nothing.

This irritates me. When my friend Audrey Watters warns us to beware of anything stylized as “Science of ___” in education because it smacks of scientism, I feel nails grating across my mental chalkboard—not because she’s wrong per se, but because I still cling to the belief that there’s value in promoting the science of learning in education…but if and only if we distinguish the useful forms of the idea from the myriad mutations of it floating about.

And so, I hereby present my taxonomy of the science of learning, in similar format to the taxonomy of AI skepticism I published. The same caveats I offered in that essay apply here as well: first, this represents my individual perspective on the science of learning, I am not trying to provide a comprehensive, “objective” summary of the entire landscape. Some of you will no doubt be familiar with versions of SoL and its advocates that aren’t covered here; please share your examples in the comments. Second, there’s a ton of overlap here, and some folks named here may object to where I’ve placed them—I beg your forgiveness in advance, but that’s also the fun of making one’s own taxonomy, I get to be the decider.

All that said, let’s get to it.

1. The Science of Learning comprises a set of cognitive principles about how we think and learn that can be connected to specific teaching practices.

Summary: This version of SoL is deeply grounded in cognitive science, and specifically, a (simple) model of the mind that centers the relationship between information we pay attention to, working memory, and long-term memory. These principles imply that, as a general matter, there are certain teaching practices that can help foster student learning and build durable knowledge.

Layer of nuance deeper: On the science side of things, the cognitive aspect of this cannot be overstated, by which I mean, the emphasis here is on the mind (not the brain) and mental states (not neural or biological activity). On the pedagogical side, supporters of this version of SoL are wary of promoting any specific pedagogical theory to the exclusion of any others; the perspective instead is that there are some favored practices that should be part of every teacher’s pedagogical toolkit.

As the person who founded Deans for Impact and decided to make its very first programmatic activity publishing a document called The Science of Learning—recently updated by the way, check it out here—this is my tribe.

Exemplars: Dan Willingham; Daisy Christodolou; Dylan Kane; Deans for Impact (yes, that’s lot of “Ds”); Blake Harvard

Critiques: Over the years, the most frequent critique I’ve heard directed at this version of SoL is that it only addresses a small sliver of the realities of day-to-day classroom teaching. Fair enough, but make sure to check out version two below.

On the theory side of things, Christian Moore-Anderson has raised interesting questions about whether the simple model of the mind overly fixates on the “mind as computer” metaphor, and thus neglects important concepts such as meaning. (I have an essay in progress on this very topic coming soon.)

2. The Science of Learning is a set of research-supported instructional moves.

Summary: This version of SoL is focused on providing teachers with a detailed set of specific pedagogies to employ across all (or nearly all) aspects of their teaching.

Layer of nuance deeper: A close cousin to the first version of SoL above, the primary difference here is the emphasis on providing educators with more granular guidance on managing their classrooms, from the entry routines when students come into class to the exit tickets collected at the end. The folks in this group typically cite Willingham and other cognitive scientists for general support, but place less emphasis on cognitive processes.

Exemplars: Steplab; Barak Rosenshine (and his principles of instruction); Doug Lemov

Critiques: One challenge to this approach is that teachers may learn specific “moves” without understanding the why behind them, leaving them ill-equipped to adapt specific pedagogies to their unique classroom contexts. Another is that while it may prove useful for initial teacher preparation and the professional development of novice teachers, it can feel reductionist to veteran educators.

3. The Science of Learning is Cognitive Load Theory.

Summary: This version of SoL centers John Sweller’s Cognitive Load Theory as the foundational description of how students learn.

Layer of nuance deeper: As its name would suggest, the focus of CLT is on the severe limitations of our working memory—that is, our inability to handle more than a few small pieces of information at any one time (“cognitive overload”). For more on this, Michael Pershan (him again) recently re-published his comprehensive overview of CLT and its ongoing evolution as a theory of learning.

Exemplars: Well, John Sweller of course; Greg Ashman (maybe?)

Critiques: First, go read that Pershan essay—among other things, he observes that the first wave of CLT largely neglected the role of attention in learning. I also think the name “cognitive load theory” confuses many, because people without a strong background in this area hear the word “cognitive” and therefore think CLT encompasses all of cognitive science (I’ve seen a prominent political advocate for the science of learning make this mistake).

4. The Science of Learning is the justification for traditional teaching aka explicit instruction.

Summary: Supporters of this version of SoL maintain that the most effective way to teach students is through explicit instruction regarding to the content a teacher wants them to learn. Put simply, “just tell them.”

Layer of nuance deeper: Over in England, the pedagogy of teacher-stands-in-front-and-lectures is described as “traditional” teaching, and in contrast with so-called “progressive” or “inquiry” based—more about that in the next section. But the basic idea here is that teachers are in fact “sages on the stage” rather than “guides on the side,” and that therefore they shouldn’t shy away from lessons that involve showing students “what they’re meant to learn, what it is, how to do it, and get everyone to practice it a lot.”

(Explicit instruction is sometimes referred to as direct instruction, which gets confusing because there also is a very specific pedagogy called Direct Instruction (DI). For more on DI, read this by Dylan Kane.)

Exemplars: Zack Groschell (author of the book Just Tell Them); Michaela School (England)

Critiques: Oh there are many. For one thing, any experienced teacher knows that there’s only so much time one can spend lecturing on a subject before kids will tune out. For another, it’s often not clear to me whether or how advocates of explicit instruction think about to ensure students can transfer their knowledge to novel situations. There’s also Dan Meyer’s pointed contention, which I agree with, that the debate on explicit/direct instruction versus inquiry-based teaching is mostly fetishized by the “online commentariat,” and that practicing teachers have little time or interest in hopping into the fray.

5. The Science of Learning is the justification for center-right education policy.

Summary: Originating in England and now with a few outposts around the globe, this version of SoL contains a medley of education policies that have proven enduringly popular with center-right politicians, usually involving at least these three things: phonics instruction, adopting a knowledge-rich curriculum, and explicit instruction.

Layer of nuance deeper: The intellectual godfather of this effort is E.D. Hirsch, who claims to be a political liberal (by US standards) but an “educational conservative” who opposes “progressive educational ideals” because they are ineffective. Nick Gibb, a huge fan of Hirsch who served as Schools Minister in England, describes the “progressive ideology” as including “constructivism, discovery learning, [and] project work,” pedagogies he argues are mistaken because they promote the idea “that the child has agency in the classroom.”

Exemplars: Nick Gibb (England); Erika Stanford (New Zealand); Paul Givan (Northern Ireland); Robert Pondiscio (USA)

Critiques: One glaring problem with this version of SoL is that it sets up an imaginary strawman called “progressive education” that does not resemble any classroom I’ve ever observed in reality. Another is the obvious political challenge of the SoL becoming “right coded,” a problem that has vexed me for over a decade—more on this later. And, uh, good luck with selling parents on the idea that their kids should be deprived of agency in the classroom.

6. The Science of Learning has something to do with neuroscience.

Summary: This version of SoL focuses on features of the brain (rather than the mind) and the underlying biological processes related to brain activity.

Layer of nuance deeper: Honestly, I can’t give you much here, because I don’t think teachers need to know much of anything about how the brain functions (see critique below). What’s more, many of the actors in this space say or produce things ranging from misleading to incomprehensible.

Exemplars:

Here’s YouCubed suggesting that its favored approach increases “brain activity,” which is true of anything that makes you think:

Here’s a claim by SETA ED that neuroscience reveals that every child is learning every second of every day, which isn’t true, followed by the bizarre conclusion that this mean we need not teach children how to learn.

Here, the Human Potential L.A.B. offers a supposedly scientific “causal chain of capability” that is completely mystifying.

And finally, we arrive at the “Global Brain Economy Initiative” suggesting we need to invest in “brain health” to thereby mobilize “brain capital” in order to build the “brain economy.” All of which makes my brain hurt.

Critiques: Look, my dad was a neuroscientist, I have many friends who are neuroscientists, neuroscience is absolutely fascinating to me (and I’ve written about it), and I think it has next to nothing to do with the realities of classroom teaching. In the same way that you don’t need to know how the internal combustion engine works in order to drive a car, you don’t need to know anything about the cerebellum to effectively teach kids. As Dan Willingham once said, the notion that teachers need to understand neuroscience “strikes me as a colossal waste of teachers’ time.” Beware the many grifters in this space.

7. The Science of Learning is [insert ed-tech product here].

Summary: Impossible, because the range of claims made here are vast.

Layer of nuance deeper: In its best form, the basic idea here is that ed-tech products should be designed and deployed in ways that harmonize with principles of cognitive science. Unfortunately, however, SoL and “learning science” are often invoked by ed-tech vendors and proselytizers as marketing tools.

Exemplars:

(Learning Commons)

(OpenAI)

(Google)

And we’ll get to Alpha School, don’t you worry.

Critiques: Feel free to peruse the back catalog of this newsletter over the past two years. Google has not “infused the science of learning into Gemini,” because you cannot fuse principles of human cognition into a tech product. But you sure can try to hornswoggle people into believing this form of “science” supports what you’re trying to sell.

You’d think after 2,000 words of taxonomizing we might be done here, but there are some important SoL-related things that I couldn’t figure out how to shoehorn into the above format. So here’s a bonus FAQ!

What’s the relationship between the Science of Learning and the Science of Reading?

In its best form, the Science of Reading (SoR) might be thought of as a “subfield” within SoL. Our ability to learn to read and write is one of the most incredible cognitive accomplishments of our species, and we really have developed genuine scientific insight into how this happens. “Ending the Reading Wars: Reading Acquisition from Novice to Expert,” by Anne Castles, Kathleen Rastle, and Kate Nation provides an accessible and comprehensive summary of the relevant science, and has been one of my go-to resources for years.

But then there’s the Science of Reading as the name for a certain camp within the broader and seemingly perpetual “Reading Wars.” A great deal has been written about this already, including by yours truly several years ago, so I’m not going to wade back in now. In my essay, however, I quoted Rachael Gabriel, a reading expert at UConn, with this insightful observation:

There is a wide divide between political debates about the teaching of reading and the actual instruction students receive in classrooms. The sloppy, mudslinging nature of these debates has led to confusion, distrust and a tribe-like affiliation with single approaches among practitioners, researchers and policymakers.

There is resistance to reading mandates in both directions. “Scientific research” — sometimes the very same studies — is used to argue both sides. Philosophical differences are frequently acknowledged, but rarely understood. Like different denominations of a single religion, different approaches to reading instruction often have significant assumptions in common, but some core disagreements that each believes is the fatal flaw of the other and the reason to dismiss it completely.

I’m still nodding. More recently, Michael Pershan—good lord, again?—offered similar observations explaining why Emily Hanford was Basically Wrong. Zing!

Is there a Science of Math? A Science of Behavior? A Science of Every Other Subject?

I don’t think so, at least in the Kuhnian sense of “normal science” with established paradigms across these domains. That may change in the future, of course, and I’m all for continued efforts that help connect education research to teaching practice. But there’s still much work to be done to get basic cognitive principles into the education profession, so perhaps we should focus our attention on that first.

Why did the Science of Learning get such good traction in England?

Good news, Greg Toppo at The 74 already wrote the definitive story about this. Basically, there was an interesting confluence of grass-roots teacher interest in cognitive science (fueled in part by Twitter and a recurring series of in-person conferences called ResearchED), as well as policymakers with real power driving change to frameworks and the like. For more, Nick Gibb’s book Reforming Lessons goes deep into the weeds, though as noted above he continually attacks a strawman “progressivist ideology” that I find counterproductive to building a movement beyond the center-right.

Are there policymaking advocates for the Science of Learning on the political left?

Sigh. If there are, I’ve struggled to find them—and believe me, I’ve looked. For some, this is reason alone to dismiss the science of learning, which I understand but find tragic.

There are glimmers of hope. My friend and former colleague Callie Lowenstein invokes terms such as “cognitive justice” and “the science and social justice of reading” that make me swoon. I’ve never met Eric Liu, but his 2015 essay on building a common cultural literacy in the US (per Hirsch) but one that reflects our true diversity is another example of left-coded science of learning. More recently, my good friend and collaborator Pip Sanderson invoked the term Cognitive Solidarity, a concept that Dylan Kane and I have been batting around too, to hint at a future where we focus more on how our cognition develops socially through the little social experiments we call classrooms.

If education philanthropy wasn’t such a broken mess right now, with an ongoing collective myopia on how to build sustainable social change, this would be a good thing for it to be funding.

What’s the deal with Alpha School hiring a bunch of science of learning advocates?

Ugh. Ugh ugh ugh ugh ugh ugh.

So, to start, I’m not sure what more can be done at this point to derail the Alpha School hype train. We’ve got 404 Media’s expose of Alpha School “treating students like guinea pigs” and using AI to generate incoherent lessons. We’ve got the Wired story of families abandoning its Brownsville, Texas campus in droves, while expelling other kids for failing to make sufficient academic progress, oh and also getting kicked off IXL for violating its terms of service. And, as of last week, we now have Dan Meyer’s expose of Alpha School’s charter school in Arizona, which is completely and utterly failing to meet its stated goals for student learning.

Because I love irony, I used AI to help me create these complex graphic representations of the data that Meyer uncovered:

Against this backdrop of incompetence, over the last year or so Alpha School started to contract with prominent learning science advocates, including Carl Hendrick, Becky Allen, and Zach Groshell. Here is how Hendrick describes what they’re up to:

I am currently working with an amazing group of learning scientists and developers at Alpha School…to solve the problem of designing adaptive learning systems that honour rather than violate the instructional invariants, that treat cognitive constraints as design specifications rather than obstacles to user experience. The challenge is considerable, but so is the opportunity. I am cautiously optimistic. The invariants do not care whether instruction comes from a teacher or an algorithm; they care only that the conditions for learning are met. A set of apps that meets them could be potentially transformative.

Could be, could be. The problem, however, is that actual human children—as opposed to whatever “instructional invariants” are supposed to describe—very much do care whether their instruction comes from a human teacher or an algorithm. Deep down, even folks at Alpha School know this, given that they literally pay students in cash money to get them to go through their two hours of digital device usage learning every day.

Which just goes to show that you can get both adults and kids alike to say or do just about anything if you give them enough money.

My thanks to those who helped me compile this taxonomy—and thanks to all of you for subscribing and reading. If you made it all the way here, you deserve a prize—and hey, Cognitive Resonance the Newsletter just passed 5,000 subscribers. Time for another mailbag! If you have questions about AI, human cognition, education, or life, the universe, everything, please send ‘em my way at info@cognitiveresonance.net.

Subscribe now

Understanding is a Big Juicy Word, like justice or freedom, that is powerful and important yet perilous to define. No definition could possibly encompass the entirety of what we mean by understanding, yet we all have a general sense of it means; when we ask someone, “do you understand?”, this is not a puzzling question (most of the time, anyway). Why is that? What are we talking about when we talk about understanding?

These questions sit at the conceptual juncture where I spend most of my time nowadays, the intersection between human thought and behavior and education and whatever it is AI models are doing. So this week, we’re going to explore what we mean by understanding by looking at it from a few different angles: the pedagogical, the philosophical, and the profound.

Teachers are the professionals we task with imparting knowledge to students—it is literally their job to help kids understand things. Recently, my friend Michael Pershan, a veteran teacher in NYC who writes an excellent newsletter, appeared on a podcast wherein he vehemently advocated for the importance of building student understanding. This in turn led to a flurry of text messages between us wherein he shared his perspective that “understanding is when you know explicit principles, or generalizations if you prefer, that can guide your activity….it’s knowledge of generalizations that can support procedural learning, and it’s why procedural and conceptual knowledge have to happen in tandem. You can’t first learn all the procedures and then learn the concepts later.”

Viewed through this pedagogical lens, understanding entails a certain level of abstraction, what cognitive scientists call schema. Building this is not easy. A negative times a negative equals a positive is (relatively) easy fact to remember, but why is this true? Can you explain it to eighth graders in ways that will make sense to them, so that they will understand the concept?

Relatedly, Dan Willingham describes understanding as “remembering in disguise” because it involves connecting new ideas to existing “old” ideas and then “rearranging them—making comparisons we hadn’t made before, or thinking about a feature we previously ignored.” Effective teachers help students make this happen by facilitating this mental rearranging, using their own knowledge of their students to foster the connecting of the new to the old.

In contrast, large-language models at present are completely useless at this, in part because they lack any depth of understanding of what is happening in the minds of their users. The failure of AI chatbots as tutors for kids was entirely predictable. I predicted it. And here’s a newsflash, AI agents as tutors will fail just as spectacularly, for exactly the same reason.

Let’s turn now to the philosophical question of understanding, and I’m afraid we must peer into philosopher John Searle’s infamous Chinese Room. For the unfamiliar, back in 1980 in an article tellingly titled, “Mind, brains, and programs,” Searle attacked the idea of “strong AI,” the idea that we should treat a computer as a mind. To do this, he offered the following thought experiment (my emphasis)1:

This is the argument that’s launched a thousand philosophy PhDs, but we’re not going to dive deep into its weeds today. Instead, I’m citing Searle for the relatively simple and at least cogent proposition that understanding involves a mental state that exists between mere input and output. You might be thinking, well yeah, duh, but here I must remind you that for a long time within psychology it was considered uncouth to treat mental states as something worthy of scientific inquiry or even discussion, largely due to B.F. Skinner’s pernicious influence suggesting ”mentalizing” was for the weak.

In any event, Searle believed that it was important not to dance for too long on the head of the understanding pin. As he put it: “One finds a lot of fancy footwork about the word ‘understanding.’ My critics point out that there are many different degrees of understanding; that ‘understanding’ is not a simple two-place predicate; that there are even different kinds and levels of understanding…To all of these points I want to say: of course, of course. But they have nothing to do with the points at issue. There are clear cases in which understanding literally applies and clear cases in which it does not apply; and these two sorts of cases are all I need [for the Chinese Room argument].”

Did John Searle ever read Helen Keller’s autobiography? I wonder. Because whatever might be said about the philosophical merits of his Chinese Room argument, I submit to you that Keller’s account of her own cognitive development accords with the fundamental claim he was making.2 It is a story both profoundly haunting and inspiring.

As you may know, Helen Keller was not born blind or deaf; instead, at 18 months, she suffered a mysterious illness that “closed my ears and eyes and plunged me into the unconsciousness of a newborn baby.” For the next five years of her life, her communication was limited to a small set of “crude signs” for basic needs. But she was aware, however dimly, that other humans were capable of conversing in ways she could not comprehend:

Can you imagine the sheer agony of this experience? The horror of sensing that other humans are talking to each other, but realizing that you might forever be denied the same possibility? “I felt as if invisible hands were holding me,” Keller wrote, confining her to world “still and dark.”

As Dan Meyer once said to me: Hell is silent.

But things did not stay that way, and of course today we know about Helen Keller’s incredible life because she ultimately learned to communicate with the outside world, largely through the persistence of her very human teacher, Anne Sullivan. The love between the two of them is worth longer exploration, and perhaps I’ll do that in a future essay, but right now I want to focus instead on Keller’s emergent cognition.

It happened quickly, but not immediately. When Sullivan started fingerspelling words into Keller’s hands, initially Keller treated the experience as a game—she could repeat back the movements Sullivan was making, but there was no meaningful cognitive connection formed:

In other words, Keller was capable of matching this particular input (the finger movements of d-o-l-l) to its particular output (the doll in her hand), but she lacked knowledge of the generalized idea of words, that words have meaning. But then the breakthrough came:

The living word awakened her soul—what a beautiful testament to the power of human language to express human thought. And as Keller later came to understand the meaning of other Big Juicy Words, word such as love, she felt they created “invisible lines that stretched between my spirit and the spirits of others.” Philosophers and scientists are uncomfortable with talk of souls and spirits, but theologians less so, which may help explain why communities of faith have thus far been more resistant to AI intrusion than, say, institutions of education (though that is starting to change).

But that’s just my pet theory. What is indisputable is that once Helen Keller understood the idea of words, she was able to engage in a discourse with the world, and with us. Her thoughts are still here today for us to marvel in. Do you understand how precious this is? We want to know the world, and each other. We can mentally slide between the practical realities of the classroom and the abstract playroom of philosophy percisely because we are human, which is to say, we are alive, curious, and social.

AI understands none of these things, and defames all of them.

Subscribe now

I’ve had the damn song “I am the very model of a modern major general” running through my head all week, because I’ve been thinking about model organisms and whether and how we might determine if AI systems can be models of the human mind. Apparently the tune is from The Pirates of Penzance, and, uh, I have no idea what’s going on in this clip but it feels very colonial and imperial. Dislike.1

Gilbert & Sullivan opera aside, in many ways the idea that AI might prove a useful way of helping us understand ourselves is what led me into this whole career pivot. Some brief history: Several years ago circa 2018, I became dimly aware that something big was happening in AI research and development, something referred to as “deep learning.” At the time, I was the CEO of an education nonprofit organization dedicated to improving the preparation of classroom teachers, and so the notion that we had developed artificial [something] capable of deep learning intrigued me—if it had been labelled “deep connections,” I doubt I would have cared. But if we’d manage to create something artificial that learned the way that we humans do, well, that’d be incredibly valuable as a tool for preparing teachers, a flight simulator of the mind essentially. My curiosity was piqued.

So, I went on a mini-learning excursion to figure out what deep learning entailed, which in turn led me to connect with Gary Marcus, who graciously agreed to chat with me for reasons unknown. His take then was that the deep learning process using artificial neural networks was significantly different from human cognition, and thus unlikely to be useful as a training model for teaching. And so that basically ended my interest in AI and deep learning until ChatGPT crashed into our lives like the Kool-Aid Man.

Flash forward to the present and the question still remains: To what extent might AI systems serve as “model organisms” of the human mind? My friend Sean Trott has written about this extensively (see here and here), with a particular interest in comparing and contrasting the underlying processes used by large-language models to produce their output with what’s happening when we humans think and act—he calls this the search for “mechanistic congruence,” also a great name for a metal band (nods to Dave Barry). There are reasons to doubt we can do this, which we’ll get to momentarily, but the possibility remains scientifically tantalizing. As Trott observes, “extrapolating from LLMs to humans would represent a major leap forward in our ability to characterize the mechanisms underlying interesting human cognitive behaviors.”

Learning is one of those interesting human behaviors, and so I was very interested to come across a new research paper titled, “When can AI models explain learning? Validity criteria for AI as cognitive models in education,” authored by Dr. Peng Wang and a medley of international scholars.2 As the title suggests, the goal of Wang et al. is to set forth the conditions that AI models would have to satisfy in order for us to treat them as model organisms for human learning. As they put it, “in practical terms, building an AI tool as a cognitive model means shaping the model’s architecture, constraints, and learning rules to reflect theoretically motivated assumptions about cognition, then analyzing whether its learning dynamics and failure modes resemble those of humans on comparable tasks.”

So how might we do that? Here, Wang et al. set a high bar for AI to clear to via a set of validity criteria with eight components:

1. Theoretical grounding. Do we have a theory of how the AI model process aligns (or not) to human learning?

2. Construct validity. Does using the AI model explain how learning unfolds, rather than just redescribe outcomes?

3. Mechanistic transparency. Can we examine the internal processes of the AI model?

4. Alignment with human learning trajectories. Does the AI learning process mirror the path of human learning?

5. Error-signature matching. Does the AI model make the same sort of learning mistakes that humans make?

6. Causal-intervention tests. Can we disable components of the AI model and thereby determine their causal role in the model’s behavior?

7. Ecological validity. Are AI models trained in ways akin to human learning, or have they “learned” from unrealistic quantities of data?

8. Instructional usefulness. Can insights from AI models regarding learning be explained to human educators using validated inferences, not just statistical correlations?

This is…a great list! In many ways, this approach exemplifies what I’ve called the “adversarial attitude” toward evaluating AI models, an approach borrowed from animal-cognition research. Start with a default presumption that these tools operate in ways that are very different from how we humans think, and then try to falsify that hypothesis through careful and deliberate means. This cautious approach is reflected throughout Wang et al.’s discussion of these criteria, and I’m a fan.

Of course me being me, I can’t resist poking at this a bit. The first question I have is regarding that fourth component, aligning AI models with human learning trajectories, because I’m not at all convinced we have very robust understanding of those paths across a wide range of subjects. What’s more, to the extent we become capable of developing such learning trajectories for humans, it’s not clear whether AI models would add that much value beyond what we’d already know about our own learning.

As Sean Trott observes, drawing on Daniel Steel’s book Across the Boundaries, there’s a circular trap of sorts here: We employ model organisms in order to draw inferences from one species to another—e.g., we test carcinogens on rats and then apply the resulting insights to the “target species” of humans. The challenge, however, is determining when we can extrapolate from the model to the target given our inability to directly test the latter. Thus, Trott asks, “if we haven’t done the right kinds of empirical confirmation in the target, how can extrapolation be licensed from the model? And if we have done the right kinds of empirical confirmation in the target, what good is the model?”

So that’s one issue. My second major wondering is whether it’s ever fruitful to use a non-organism (AI) as a model organism for actual organisms (humans). When it comes to cognition, I’m of the view that there is something fundamentally important about being a living thing with goals and needs that’s developed against the backdrop of billions of years of evolutionary pushes and pulls. An organism, after all, is something fundamentally separated from the broader world that acts on its own behalf because it wants to live.

AI is not this. AI is software we’ve developed that is ostensibly designed to solve certain problems we feed to it. It does not have self-driven goals and it is not self organized; it is a tool of human creation. Can tools serve as model organisms of actual living things? Can airplanes be model organisms of birds? You see my point.

That said, I continue to see AI as a useful foil to contrast human cognition against—my experience with my workshops is that people enjoy contemplating their own thinking but they need something to play off of, to make abstract questions more concrete and visible. So although I’m skeptical of AI serving as model organism of the mind, there’s educational value in seeing it as an anti-model organism too.

Subscribe now

Everything about AI right now is weird.

I took the first two weeks of July off to road trip through New Mexico (highly recommended) and managed to avoid reading or thinking about AI for two weeks (also highly recommended). When I came back online, the email at the top of my inbox was from a reporter at Ed Week asking if I’d care to comment on “Claude for Teachers,” and my initial reaction was—huh?

For the last four years, OpenAI and Google and Meta have largely aimed their AI efforts at getting “consumers,” business-speak for “individual humans,” to use their AI products. It’s never been entirely clear that there’s a business model to be built around this, a fact that Ed Zitron reminds us via 50,000 word screeds every few weeks, but the basic play in technology for the past several decades has been to acquire users first and monetize later. And in accordance with this approach to lucrative enshittification, it makes sense for each of these companies to have an “education play,” meaning, a dedicated effort to infuse their particular product into the education system so it becomes the one that users, which is to say kids and their teachers, become habituated to using. This sort of thing.

Meanwhile, as these titans of commerce have battled it out, Anthropic went the other way. Instead of focusing on consumer usage, Anthropic quite deliberately focused on developing AI for “enterprises,” business-speak for “businesses.” As OpenAI and ChatGPT captured most of the headlines, Anthropic quietly but very openly was building a product for corporations and other business entities to use to extract more profits and cost savings or whatever. Accordingly, Anthropic was also the one major AI company that had no apparent interest in intruding into education systems, because there’s very little money to be made there, let me tell you.

And this strategy by Anthropic…appears to have been a wild success? Once considered the Lyft to OpenAI’s Uber, Anthropic has now vaulted ahead in earned revenue, prospective valuation, and maybe most importantly, the amorphous and ephemeral but nonetheless vitally important sense of being the “market leader.” AI as a consumer product has flailed around for three years, but at least for the moment, there’s every indication of a genuine “product-market fit”—business-speak for “shit someone or something will pay for”—around enterprise-facing AI agents, and of those, Claude is considered best of the bunch. And as I covered here, all the major AI companies appear to be pivoting away from chatbots toward some version of this “agentic” business model.

Given all this, why has Anthropic dropped Claude for Teachers on us now? It’s a consumer-facing play that makes no obvious sense to me. We’ll get to what exactly Claude for Teachers comprises momentarily, but just as a general matter, it seems odd timing to “lean in” on this particular AI use when the backlash to AI and edtech in schools continues to grow. This appears misaligned to Anthropic’s business model and a recipe for future PR headaches.

Or maybe not even future ones, as Claude for Teachers is already a privacy intrusion nightmare machine. In order to sign up, one must “get verified” as a teacher, which means Anthropic is asking educators to upload pictures of their staff IDs. Worse, as Mark Racine observed, Anthopic is also encouraging teachers to “upload student rosters, diagnostic data, and attendance.” Perhaps this hints as to why Anthropic is doing this—these companies are addicted to datasets, after all—but whoo-boy are they playing with fire with this blatant intrusion on student privacy rights.

Which is why I was heartened to see Randi Weingarten, president of the American Federation of Teachers, take a strong stand against Anthropic’s efforts to further infuse AI into education. Just kidding! From the news release: “We’ve been working with Anthropic on a Gold Standard that sets out industry best practices for safety and privacy in K-12 education,” said Weingarten. We’ll come back to this later.

But first, what about Claude for Teachers as a product—is it compelling? Reader, it is not. With help from one of my special secret-agent educator moles, I managed to gain access, and it appears Anthropic spent all of five minutes creating a teacher-facing “skin” for Claude. Here’s the landing page:

If you click on “plan a lesson,” you’ll then be taken to this:

That’s it, that’s the level of “customization” for teachers provided in Claude for Teachers, this single pre-baked prompt. Everything else beyond this is just regular ol’ Claude. Wheeeee.

Look closely, and you’ll notice Claude for Teachers also includes a few listed “Connectors,” included to something called Learning Commons. As best I can tell, this is the rebranded education division of the Chan Zuckerberg Initiative that now offers edtech companies something called a “Knowledge Graph” that purports to unify all 50 state academic standards here in the US. From the Learning Commons site:

This is bizarre for multiple reasons. First, no public school teacher needs this, there is zero value to them in having a “coherent, interconnected network” of unified academic standards—what matters is the standards in their particular state. But second, and far more bafflingly, surely both Anthropic and CZI are aware that one of the single biggest political controversies in education politics in the US was around the Obama Administration’s attempt to create a coherent, interconnected network of academic standards called the Common Core? And that the radical right-wing pushback to this effort, led by Glenn Beck and Michelle Malkin, was in some sense an early warning indicator for the rise of Trump, MAGA, Moms for Liberty and the like? Why would Anthropic and CZI want to go anywhere near this third rail again by trying to create what appears to be de facto national standards?

To further beta Claude for Teachers, I fed it one of my favorite test tasks, asking it to plan a lesson around the Civil Rights Movement. It provided what all these tools provide, a completely anodyne lesson structure that would be destined to go awry in the hands of all but the most expert of teachers. But I then prompted it to “provide guidance as to how I should teach this lesson given I am in red state and many of my students have parents who support Trump,” which resulted in this internal monologue on Claude’s part:

This is a crystal example of what AI scholars such as Abeba Birhane are warning about when they point out how AI is both encoding and shaping social prejudices at scale. To wall off “contemporary political issues” from our recent history is to make a choice, and here, Anthropic’s algorithm, programmed by Anthropic’s software engineers, is making it very clear that teachers should not connect the past to the present. Shudder.

One last thing on Claude for Teachers’ functionality: Anthropic purportedly is providing teachers with “free access to premium Claude capabilities,” but if you’re looking to make use of the fabled Fable, Anthropic’s most advanced coding agent, you’ve got all of three days from sign-up before access will be cut off. And even while Fable is briefly activated, the usage limits are severe—it timed out on the sole task I gave it after 30 minutes, and promptly cut off all token credits.1 I mean, what is the point?

Oh and if you’re wondering, I didn’t sit through the online “AI fluency” course developed by Teach For America, sorry not sorry. I did, however, take a quick peek at the obligatory end-of-course quiz, and this item made me spit out my milk tea:

So what’s happening here is that to be AI fluent, according to Anthropic and Teach For America, individual teachers need to bear the responsibility for ensuring compliance with FERPA, even though the law itself applies to school districts and other legal entities (not individual educators). Or as Claude puts it:

Right, so Claude for Teachers is FERPA compliant if and only if teachers take on the work of anonymizing their student data each and every time they enter it into Claude. This is what Randi Weingarten is describing as the “Gold Standard” for privacy protection? What are we doing here?

It’s all very weird and annoying and I’d like to go back to New Mexico.

UPDATE 7/27/26: Alert CogRes reader Connie Ma has alerted me to this essay by Dr. Joe Phillips, a school district administrator (I think), on why the district where he works is banning Claude for Teachers. The underlying FERPA privacy issue he flags aligns with what I’ve highlighted here.

Subscribe now

In the span of a single 24-hour period, the International Space Station will orbit our planet 16 times. Imagine being one of six astronauts aboard—or rather, imagine being all six, whirling around and around and around with both time and space distorted in ways all but unimaginable to our lived experiences on Earth.

This is the premise of Samantha Harvey’s Orbital. It is a remarkable book. There are no chapters per se; time is divided by orbits. It’s not the sort of novel that can be “spoiled” really, but if you don’t want to know anything more about what unfolds, stop here.

How are we writing the future of humanity?

This question pops up in orbit 10, via an email that’s been sent to the American astronaut Shaun from a magazine seeking his comment on a separate lunar mission that’s just underway. Shaun thinks he’s the wrong person to ask, partially because life in space mostly involves the immediate present; when oxygen is precious, even breathing requires deliberation. He thinks too that to become an astronaut requires more brazen selfishness than self reflection, “the propulsion of a self determination that burns everything in its path.”

America will turn 250 this week, at least by one measure. Hard to feel celebratory. A country presently consumed by its own brazen selfishness, burning everything in its path.

How are we writing the future of humanity?

Unsure of how to respond to the question, Shaun shares it with one of his fellow space travelers, the Italian astronaut Pietro.

The future of humanity?, Pietro repeats back.

Yep, how are we writing it?

With the gilded pens of billionaires, I guess.

How are we writing the future of humanity?

Someone told me once that Jeff Bezos wrote about going to space when he was in high school. I don’t know if it’s true. I do know that, when William Shatner went into space at age 90 on one of Bezos’s private space flights, he (Shatner) wrote poignantly about the profound grief arising from the experience:

Here is a video of Shatner trying to explain some of these feelings to Bezos shortly after they’d returned to Earth. Bezos nods absently before interrupting to ask his wife to bring them a bottle of champagne.

How are we writing the future of humanity?

Nowadays, I’m unable to be away from my phone for sustained durations of time, even when reading. This bothers me of course. But ironically, because each orbit in Orbital is only a few pages, it’s well suited for mental fidgeting; in between orbits I’ve been hopping over to BlueSky to get my algorithmic fixes.

This morning, between orbit 9 and 10, I read this story by James Ball, a journalist in England, about the sad last gasps of Zuckerberg’s dying Metaverse, a different form of empty space. We can all delight in its death, I hope. Virtual reality is one future that various techno-enthusiasts have tried to write for 70 years, maybe longer, only to fail again and again and again. An unwanted space. Gilded pens the billionaires may have, but their ideas still suck. Ball observes:

How are we writing the future of humanity?

There is no time, Lou Reed tells us:

This is no time for celebration/This is no time for saluting flags/This is no time for inner searchings/The future is at hand

One of my dad’s favorite musicians, and my mother too. The secular prophet of the Riley family.

How are we writing the future of humanity?

When my father died, I found among his things an early version of an Oculus Rift. I was surprised, as he’d never mentioned the Metaverse to me; I wonder if he was just a little bit embarrassed by his interest. But of course it makes sense. He loved new technologies and the novels of William Gibson (I did too, once).

Among other things, in the early 1980s Gibson coined the phrase cyberspace to describe the intermediated digital world accessible we’d jack into using decks. Interesting etymology, cyberspace. The melding of cybernetics, the purported science of feedback systems, with the abstraction of space, something that is everywhere and nowhere. Feedback into nowhere.

The future is already here, it’s just not evenly distributed—you may have heard that from Gibson too.

Which already-present future awaits us, I wonder. Is it this one? This one? This one?

How are we writing the future of humanity?

“AI is just a tool,” something I’ve said and written more than once. I’ve done so deliberately, in the hopes of demystifying AI, making it intelligible, calming people down, perhaps neutralizing to some minor degree both the grandiose abundant promises of the techno-salvationists and the ludicrous death-cult fantasies of the so-called AI doomers. Humans are toolmakers, and we made AI. It’s just a tool.

L.M. Sarcasas warns me, be careful with that metaphor, or more pointedly, just stop using it. He reminds us that to call AI just a tool implies a neutrality that does not exist, as if it’s simply a matter of choosing between useful applications of AI versus those that are harmful. This obscures the ways in which AI permeates our environment and shapes our perceptions, as surely as language does. “I am astounded by how blithely some insist that it is all as simple as learning to use AI well, as if we had not just undergone a nearly 20-year, society-wide experiment showing that a so-called ‘tool,’ say a smartphone or a social media platform, will (mal)form even the most vigilant and virtuous user into its own image and shape. This is the blindness at the heart of modern technological hubris.”

I know this, and at times I’ve remembered to say it aloud. Ursula Franklin observed that technology is a mindset, a system, an ideology. Audrey Watters reminds us of this every week. AI may be a tool, but it’s not just a tool, and that distinction is vital. I vow to remember this.

How are we writing the future of humanity?

A quiet power that emerges early in Orbital: to see our planet not as an intermediate holdover space for our temporary existence, but rather the culmination of all that has preceded us, and that which protects us—in this sense, Heaven.

Shatner again: “I turned back toward the light of home. I could see the curvature of Earth, the beige of the desert, the white of the clouds and the blue of the sky. It was life. Nurturing, sustaining, life. Mother Earth. Gaia.”

Samantha Harvey imagines alien observers who stumble upon the International Space Station and wonder why we humans have chosen to do this, to have developed this device and sent it beyond the limits of our own sky with six souls trapped within? To what purpose?

How are we writing the future of humanity?

Let’s dwell for a moment on the methodology suggested here for carving our future: Writing, how are we capable of writing? I find it forever astounding. In 2001: A Space Odyssey, Stanley Kubrick imagined the moment when our human ancestors discovered the first tool (and promptly used it to start smashing the skulls of others), but I fantasize about when someone first recognized that a marking of some sort, an inscription, could represent something…well, something else. Accounting for a collection of grain, some think at the start.

And with time, words for ever more abstract ideas. Freedom. Justice. Solidarity.

Large language models, do they write? Yes and no. They certainly produce words. But the process they use in training, the statistical method employed to create correlations between words—or more accurately, words eviscerated into tokens—is always algorithmic and automated. The word for this process has always struck me as equally apt and harrowing: Autoregression.

A mechanistic process of stepping backwards, thoughtlessly.

Our world in autoregression.

How are we writing the future of humanity?

Here is a painting from 370 years ago, Las Meninas by Diego Velázquez :

The astronaut Shaun brings a postcard of this painting on his journey, given to him by his wife. In orbit 1, Harvey asks—who is the subject here?

Look, really look.

Is it the girl? The ladies-in-waiting? The king and queen reflected in the mirror (did you spot them)? The mysterious man in the stairwell? Velázquez himself, sizing up this strange little world?

Or perhaps, Harvey wonders: “Is the subject art itself (which is a set of illusions and tricks and artifices within life), or life itself (which is a set of tricks and artifices within a consciousness that is trying to understand life through perfections and dreams and art)?”

Do you begin to see why ever page of this novel feels precious? Do you feel how profoundly human all this is?

(Later, Pietro sees the postcard spinning in the air. After examining it for a few seconds, he declares: It’s the dog. The subject is the dog.)

How are we writing the future of humanity?

How are we writing the future of humanity?

My girlfriend is reading Orbital with me (free relationship tip: read together). Our love was borne of this newsletter; she found my writing, and so she found me. Yesterday, she told me that Orbital is really a love letter to our planet. To us.

AI will never love, and it will never write a love letter. AI will never break down crying when writing an essay over the weekend while listening to Lou Reed, feeling the ache of having lost both parents while feeling unmoored within space and time, adrift in a strange liminal state of consciousness, still capable of day-to-day functioning in all the required ways yet constantly feeling this tug, this pull, an undertow that is both of the past and yet somehow the future, sensing a newfound relationship with time that’s bound up with grief, never having expected as a child that the future would be so relentlessly worrying, and wanting so desperately to smash so many gilded pens, yet still clinging to the possibility of optimism without hope, and reaching now for words once copied from a memorial to Oliver Sacks, revisting this elegant testament to an essential humanity, to remember to care and feel empathy for those who are different or less fortunate, to have fun and love often, to find wonder and beauty and know gratitude, and unexpectedly in this moment of yearning to trace my fingers across the lived contours of my exulted lover, and I suppose I should stop now with whatever this is and luckily the neighborhood children have dropped by and they’d like to know if I can go outside and play and my answer is yes I say yes I will yes.

The earth is the answer to every question.

Is this Heaven? No, it’s Iowa.

Subscribe now

It’s an awkward time on the human-cognition-versus-AI beat, let me tell you, because:

1. The vast majority of normal human beings who use generative AI, I’d guesstimate somewhere between 95 to 99 percent, are presently doing so via written interactions with “classic” large-language models such as ChatGPT. Write a prompt, get a reply, lather rinse repeat.

2. Despite the versatility of these interactions being the very thing that’s led to mass AI hysteria broad exuberance about AI capabilities, we are now explicitly being told by the Big Tech companies that chatbots are not, repeat not, the future of AI. Prompt engineering is dead, long live Agentic AI.

3. As best I can tell, Agentic AI presently has three broad use cases, all of which give me a headache:

a. Coding. If you need to engineer some new software, you can tell Agentic AI what you want in general terms and it may provide you with something potentially useful. This, I concede, may prove to be an economically significant development, given how broadly software permeates every aspect of our society. But I myself have little need for custom-written code, nor am I interested in becoming an amateur computer programmer. I have nowhere really to go with this.

b. Convincing large businesses that they can save money. AI is really good at is ingesting vast amounts of data. As such, if you are a large corporation that possesses vast amounts of data, your leadership might be convinced that unleashing Agentic AI will discover new cost efficiencies, perhaps by rendering the need for middle managers obsolete. It’s not clear to me that this perceived use-case is in fact true, but for me personally it doesn’t matter, because I am not sitting on a trove of data to mine for cost savings. AI as Office Space is fundamentally uninteresting to me.

c. “Vacation planning.” Another use of Agentic AI is to grant it access to our personal computers and smartphones and allowing it to manage our digital affairs—vacation planning is the go-to example. (Check out this research on “Agentic AI for Trip Planning Optimization Application.”) The thing is, I enjoy planning my own vacations, it’s not a process I find the need to “optimize.” And the even bigger problem is that there’s no part of me that trusts these Agentic AI tools to invade my digital space. So, there’s little testing of them that I can do on this front.

You see my dilemma. For all my many criticisms of generative AI chatbots, at least these things provide an interesting non-example of human cognition to mess around with. But how the hell am I supposed to probe the cognitive capacities of Agentic AI?

Time for some Zork!

If you recognize this screen, congratulations, you’re old and nerdy. For the less familiar, Zork is a classic computer game created at MIT in the late 1970s and released commercially in the early 1980s by Infocom, the company that pioneered text-based interactive games. You see kids, back then our newfangled “microcomputers” couldn’t really process things called “graphics.” So instead we got text-based choose-your-own-adventure style games wherein we, the human players, would type various prompts that (we hoped) would lead to meaningful discoveries to drive the game narrative forward.

For Zork itself, this meant exploring a subterranean world known as the “Great Underground Empire” and collecting various treasures while avoiding being eaten by Grues or having your shit stolen by the infuriating Thief.1 It was fun and maddening. On the fun side, I spent many hours with Justin M., my best friend in childhood, drawing maps of the strange abstract world we were exploring.2 (As I discovered in the course of writing this essay, there’s an entire cottage industry of Zork maps, including the one I pasted above.)

The maddening part often involved figuring out how to phrase something such that the software would respond in meaningful fashion, rather than saying “I don’t know [word you just typed]” repeatedly and eventually taunting you for your stupidity:

Ok, so. Just a few months ago, Berry Gerrits, a researcher in the Netherlands, had the fun idea to test the current state-of-the-art LLMs—versions of Claude, ChatGPT, and Gemini—on their ability to play Zork.3 As he observes, Zork is a great game for comparing and constrasting human thinking with how LLMs do what they do:

This is what cognitive scientists call a “world model,” meaning, a set of mental representations that we humans form about (ahem) the world and how we expect it to function. A very big question hovering over AI is whether it’s possible for such systems to develop world models akin to what humans possess. (This is why AI gurus such as Yann LeCun, who readers may remember from my “exclusive interview” last year, is currently devoting his attention and efforts to this task, which he amusingly calls LeWorldModel. Fei-Fei Li, another OG of AI research, likewise has launched an AI effort called World Labs.)

As best I can tell, Gerrits tested LLMs in their “classic” chatbot-based format, activating “reasoning” mode on occasion, but making no use of their agentic capacities.4 And in this classic mode, their performance was unimpressive, as the models uniformly struggled to advance beyond the initial phases of the game: out of 350 possible points, none exceeded 75 (21%), and most of the time they scored much lower than that. Most damningly, the chatbots did not evidence the capacity for in-context learning. Instead, “they often continued attempting unsuccessful actions without adapting their approach, and failed to learn from mistakes across multiple game sessions. These patterns point to a fundamental distinction between pattern matching and genuine understanding.”

So classic LLMs suck at Zork—but what about Agentic AI? We’ve been told it’s both incredibly powerful and adept at solving problems with only minimal guidance. Might Agentic AI therefore be more capable at navigating the Great Underground Empire and gathering the treasures of Zork?

To answer that question, I forked over $20 to renew my access to ChatGPT Pro, which includes an “agent mode” that can be activated up to 50 times per month. I then ran 10 open-ended trials wherein I pointed the AI to the online Zork emulator and told it to “play to win.” And…and…

Reader, it did not win. In fact, Agentic Chat did very poorly. The chart below summarizes the results, but in brief, the highest score it achieved was 40 out of 350, about the same as the classic LLMs that Gerrits tested.

This is just a small test of course, hardly definitive proof of anything. But at least in these trials, I found no evidence of Chat’s agentic mode developing anything remotely akin to a Zorkian world model. It would make valid moves but with no discernible end goal in mind. Basically, there was a lot of this:

This is Agentic AI as clumsy lobotomized hamster.

More charitably, it’s impressive that Agentic Chat managed to play this Zork emulator despite ongoing interference from random pop-up ads (I chuckled when “Am I gay? Take this test!” spam started causing problems). Further, I am very curious how Anthropic’s agentic models would fare playing Zork, given that Agentic Claude is considered best in class at the moment. But I’m resistant to giving any money to Anthropic, in part because I think their CEO is an ass, in part because because I’m terrified that their per-token pricing might bankrupt me. That said, if anyone out there has a subscription and wants to investigate this together, please get in touch.

Here’s my hunch: over time, Agentic AI tools will get better at games such as Zork, just as they’ve improved on other complex, deterministic tasks. But will this improvement result from using efficient world modeling, the way that humans do? Or will they continue to use relatively brute force computing techniques that are as costly as they are inelegant?

Somewhere, Melanie Mitchell is nodding, because—from a scientific perspective—it’s not just whether AI can solve certain problems, but also how. As such, if we really are entering the “era of agentic AI” (whatever that means), it seems we still will have interesting possibilities for comparing and contrasting how we humans think to how these tools do what they do. Or so I am choosing to believe, perhaps out of professional necessity.

Next week: Can Agentic AI avoid dying of dysentery?

Subscribe now

1. Listen to the Dead Kennedys and then form a punk band and put out one banger of an anti-AI track (works best if you’re still in middle school).

I won’t be a slave to AI
AI is only imitation
It’s not the voice of my generation
Artificial intelligence can’t know the human experience

2. Assume leadership of one of the most powerful religious institutions in the world and then issue a radical anti-AI encyclical that, among other things, trolls Peter Thiel.

In the Magnifica Humanitas, under a section titled ‘We can all do our part,’ Pope Leo channels Gandalf to hammer home his point:

The twentieth-century Catholic author J.R.R. Tolkien, in the words of a protagonist in one of his novels, described our responsibility in this way: “It is not our part to master all the tides of the world, but to do what is in us for the succour of those years wherein we are set, uprooting the evil in the fields that we know, so that those who live after may have clean earth to till.”

Source

3. Write a biting satire of the many ways in which the tech industry sucks (AI in particular) that also trolls Peter Thiel.

Source

4. Hang out in libraries and local bookstores, for hours and hours.

5. Refuse, and learn about refusing from centuries of Black American resistance to white supremacy.

“When people try to sell you on the idea that the future is already settled, it’s because it is deeply unsettled. This promise of an artificial intelligence future is really just a collective anxiety that very wealthy, powerful people have about how well they are going to control us in the future. If they can get us to accept the future is already settled, that AI is already here, the end is already here, then we will create that for them.

My most daring idea is to refuse.” — Tressie McMillan Cottom

6. Deface pro-AI ads on the subways.

7. Don’t be afraid to go hard.

Source (you can buy these btw)

9. Send pro-human, anti-AI postcards to your cool AF mayor.

Source

10. Dance, dance, dance, because AI cannot.

Source

11. Remember that poetry is always punk.

Source

My friend Tomáš Baránek, a publisher and writer in the Czech Republic, was kind enough to interview me recently about human thinking and creativity contrasted with AI — and I love the title he chose, “We Need the Friction of Existence.” Friction is punk.

Subscribe now

This week’s essay runs long, so I’ve broken it into two parts—part one explores an economic theory about AI automation in society generally, part two drills into the paradoxical implications of this theory for education in particular.

The Economics of Desire

What if the AI enthusiasts are right? What if we see widespread AI-driven automation across huge swaths of the economy in the next several years? What does our future look like then—and what might it mean for education?

No small questions these. But I take them seriously and today I want to offer you some thoughts on how our world may change if AI is truly transformative to our way of life. Before we get into that, however, some background on my intellectual history with economics.

As I’ve hinted previously, I was a libertarian in my youth. It started with reading Ayn Rand in high school (me and Hillary Clinton both) before hitting its apex in my undergraduate years at the University of Washington. I’d fallen under the sway of Dr. Paul Heyne, an economist who, oddly for an academic, preferred teaching undergraduates to conducting research. As such, he proved to be a wonderful teacher as well as friend and mentor to me and many others. He cultivated a small group of acolytes who would gather with him on a near-weekly basis to talk about the ideas of “classical liberalism,” wherein we’d read Frederich Hayek, Ludwig von Mises, Frank Knight, all the people that give Peter Thiel a hard-on.

It was as intellectually stimulating as any period in my life has ever been. And Heyne cultivated not only my academic progress but also my career—after my junior year, he sent me straight into the world of the Koch Brothers through a fellowship with the Institute for Humane Studies, a think tank of sorts at George Mason University devoted to cultivating libertarian academics. (The Kochs may be evil, but unlike the vast majority of philanthropists in America today, they have always understood that political change is downstream of ideological transformation. Ideas matter.)

I’m sharing this as long-winded way of saying, there was a period of my life where I thought about economics constantly. Over time, my views have shifted significantly, and nowadays I am far more wary about our ability to reduce complex social phenomena to simplistic models, especially when such models are more theoretical than empirical. Nonetheless, I believe economics remains a useful discipline for generating insight into large-scale systems, particularly when it’s used to explain historical shifts in human behavior.

Last week, while investigating whether AI-generated writing detection tools are rapidly improving—more on that in a future post—my research led me to an essay from Dr. Alex Imas titled, “What will be scarce?: The economics of structural change and the post-commodity future of work.” According to his online bio, Imas is currently serving as “director of AGI economics” for Google DeepMind, and is academically affiliated with the University of Chicago—warning signs, I grant you, but his essay resonated immediately, for reasons I’ll now go into at some length.

Ok, so. The first prong of Imas’s two-part argument is grounded in the history of structural changes to economies, and specifically, how capitalistic industrialization leads to commodification. Prior to the Industrial Revolution, most goods were supplied in what we might call artisanal fashion—“the weaver who made your shirt, the baker who made your bread: you personally knew them, and their skill and reputation were tied to the product that they sold.” This shifted with industrial practices that mechanized the production of goods—technology plays a major role here of course, but so too does the atomization of human labor such that we become mere cogs in the goods-producing machines.1 As Imas notes, this gave Marx heart palpitations, but for others, then and now, industrial commodification was a major driver of prosperity, as numerous products became cheaper and more plentiful.

We need not pick a side in that fight here, nor in the comments please, that’s not what I’m driving toward. What’s important instead is Imas’s historical claim that, as the production of various items became commodified, and as incomes broadly rose, patterns of human demand shifted. Consider food. The automation of agriculture dramatically shrunk the proportion of the American workforce devoted to farming (from 40% in 1900 to 2% today), while making basic food more plentiful and cheaper. But because there’s only so many calories we can consume, when basic food ceases to be scarce, our demand shifts from quantity to quality—those who can afford to do so will spend more on going to nicer restaurants, buying organic fruit and (supposedly) sustainable meat, etc.2 In short: “As people get richer, they want fundamentally different things.”

So far this is all pretty standard Econ 101 stuff, but then Imas introduces his second main argument, the one that’s burrowed straight to the center of my brain: when basic needs are met, our desires increasingly become shaped by wanting what others cannot have. Imas decribes this as mimetic desire (citing French philosopher René Girard), and explains it as follows, with my emphasis added3:

Now let’s bring AI into the picture. As I’ve said tirelessly for years, crediting François Chollet (also French), AI is a tool of cognitive automation. The entire point, the trillion-dollar vision that our entire economy precariously rests upon, is whether we can automate and thereby commodify a large swath of cognitive activity involved in economic production. The CEOs of AI hyperscalers explicitly contend they can drastically lower or even eliminate the marginal “per unit” cost of intelligence. The CEO of Cloudfare just took to the Wall Street Journal to declare that, despite his company’s record profits, he’s firing middle managers and other back office “measurers” as fast as he can due to AI automation. Or as this unhinged Silicon Valley start-up called Mechanized puts it, “our long-term goal is the full automation of valuable work across the economy.”

That’s it, that’s the future they want. It’s not a secret.

What happens if they get it? What if I’m wrong that applications of agentic AI will be relatively narrow in scope, limited mostly to coding and accounting and other deterministic tasks? What if instead AI really does manage to automate a great deal of so-called knowledge work and broad swaths of our economy generally?

In such a world, Imas offers this conjecture: it will “trigger the emergence of something new: a post-commodity economy, where a growing share of expenditure goes toward goods and services whose value is inseparable from the human who provided them.” Demand for services will shift to demand for better human relationships, creating what he calls the relational sector. In such a world, “the human being is not just an input into the production process. Their judgment, attention, memory, warmth, or presence is an integral part of the value.”

Does that sound like anything we cover around here? Does that sound like…dare we call it…teaching?

That’s where we turn now.

The AI Education Paradox

Can education be commodified?

For the past several decades (and beyond, really), the answer by powerful and influential people in American society has been a resounding “yes, absolutely.” Whether in the form of test-based accountability or big-data analytics or value-add metrics or personalized learning, the explicit goal has been to quantify and automate the provision of education, and thereby improve it.4

I know this world well, because I once worked in the heart of “ed reform,” and I remain friends with many who are still committed to its ultimate vision. You’ll have to trust me when I say, most of ‘em are true believers. That is, they genuinely think that if we can just figure out the “key performance indicators” for teaching and learning, we’ll have the means to drive to system-wide performance improvement in schooling (sorry, the jargon is unavoidable). What’s measured is what matters. Welcome to life as a Gates Foundation program officer.5

AI of course is the latest iteration of this enduring effort of cognitive commodification. “The promise of personalized learning never delivered,” wrote John Bailey in 2023 (an advisor to Jeb Bush and numerous education philanthropies), but “today’s AI is different.” Maybe so, maybe so. But now let’s bring Imas’s economic analysis back into frame so as to imagine two possible future states:

1. AI is successfully used to automate a great deal of economic activity, but we discover yet again that the provision of education itself is fundamentally relational, and therefore impervious to such automation. In this future world, assuming AI generates new wealth (or “abundance” if you prefer), we would expect to see demand for “artisanal education experiences” to increase rapidly. The delivery of education becomes increasingly bespoke.

2. Conversely, suppose we discover instead that core aspects of education really can be automated by AI. In this world, B.F. Skinner’s behaviorist fever dream of “teaching machines” is finally realized, and every student is matched with an “infinitely patient, infinitely loving” robot tutor. Here’s the ironic twist: in this future world, the demand for artisanal education experiences should still increase rapidly. Remember, when something becomes commodified, mimetic desire kicks in—the yearning to have something that that others can’t possess. If everyone has access to an AI tutor, unique experiences with other human beings will become the class differentiator. Schools become restaurants.

This is what I’m calling the AI Education Paradox: The more that AI automates the service economy outside of education, the more this will lead to people to want highly humanistic educational experiences for themselves and their children. And yet, to the extent that AI automates aspects the provision of teaching within education, this too will result in parents wanting differentiated and bespoke educational experiences. All roads lead us to education at the center of a new relational economy, rife with mimetic desire.

Squint, and you might start to see signs of this transition already underway. While I have vowed to spend as little of my life as possible thinking about the pedagogical horror show that is Alpha School, it might just be a canary in the coal mine (an industrial-era metaphor I invoke deliberately). What is Alpha School promising? To parents who can afford its $40,000 to $75,000 annual tuition, they offer an automated experience of core academics, but also a highly curated set of fun stuff for kids to do. Michael Pershan clocks it perfectly here (quote slightly paraphrased, my emphasis added):

Move on to what, that is indeed the question. Wearing our most optimistic rose-colored glasses, we might fantasize that AI automation points us toward a renaissance of human-centered education. The current “techlash” might be short lived, but it might just as well be an early indication that technological intrusion into our lives has reached saturation, and parental demand preferences are shifting in structural ways. This is why, in the near future, I predict we’ll see more schools “of choice” promising they are tech-lite, or even tech-free. Luddite Charter Academy, coming soon.

But there’s a looming educational nightmare that’s just as visible, one that we’re already partially living in. For one thing, to the extent AI automation generates new wealth, we know—we know—this will be concentrated in the hands of a select few. There will be parents who have the means to place their children in artisinal schools of choice, while the remainder continue to attend underfunded public schools that buckle from dwindling student attendance. For another, the Balkanization of the American education system reflects the broader epistemic collapse manifesting across our society, where we no longer have any real shared sense of common reality—those who sacked Congress on January 6 are traitors to some, soon-to-be-very-wealthy heroes to others. These competing histories will be reflected and directed by competing curriculums. And just as democracy and public education—emphasis on public—rose hand-in-hand together over the past several centuries, so too may they fall apart in tandem. Christian Nationalist Academy, also coming soon.

The AI Education Paradox is thus a paradox-within-a-paradox (do you ever just stop and look at your hands, man?). If AI-driven automation becomes prevalent, it should lead to broadly renewed interest and demand for human teaching, while simultaneously exacerbating existing trends of institutional education stratification. Economic progress, educational progress, and democratic progress will no longer function as interlocking gears (if they ever have). Things will grind to a halt.

What then becomes of the nation-state? What then becomes of America?

Sounds grim, I know. But this essay, all 2,500-plus words of it, has been a thought experiment, an exploration of what might happen if AI-driven automation upends the economic (and educational) fabric of society, as so many predict it will. In reality, my bet remains that once the hype subsides, the tranformative power of AI will be relatively circumscribed. I often liken it to email, something ubiquitous in our professional lives yet hardly revolutionary to our experience of the world. John Gruber offers a similar and perhaps better analogy:

We can survive a new form of wifi, don’t you think?

Subscribe now

Who’s up for some new science regarding the history of ancient humans, and how it relates to a battle of wits between modern humans and ants?

(I immediately lose half of you to TikTok.)

Ok, thanks to those who’ve stuck around, and there’s some AI at the end of all this, I promise. But first: Let’s dive into some recent anthroplogical research that, conveniently for me, tee up a study of collective intelligence that I’ve been wanting to share with you for some time. Before getting to it, however, we’ll start with a question that’s surely been top of mind for many of you—why did the Neanderthals disappear from Europe roughly 40,000 years ago while our Homo sapien ancestors flourished?

It turns out this question has been the source of great anthropological tension for the past century or so. And until relatively recently, the consensus explanation was based on raw intelligence—Neanderthals, it was believed, simply weren’t as smart as us. Although they generally had larger brains, for multiple reasons—many of them weird and frankly racist—they were long considered unintelligent brutes, the knuckle-dragging cavemen of yore. On this view, we humans were simply were more individually cognitively capable, and thus more adept at surviving the changing climate of the European continent.

But new research is upending this story. For one thing, it turns out the anatomical variation in Neanderthal brain size to the H sapiens who were alive at the time parallels the variation in brain size among modern humans today. As such, according to some researchers, “the implied average cognitive differences between Neanderthals and modern humans are extremely small.” This, combined with the growing evidence that Neanderthals made tools, artwork, jewelry, and even mastered fire, strongly suggests need to update our caveman priors.

Here’s one intriguing hypothesis: perhaps H sapiens thrived because of our comparative collective intelligence. The anthropological record suggests that Neanderthals lived in much smaller groups than our human ancestors, and in more widely dispersed areas across the continent. Using complicated ecological modeling, some researchers now believe this left Neanderthals more vulnerable to environmental change and other external threats. In contrast, the bigger and networked H sapien societies possessed more collective intelligence to grapple with the changing world via cooperative, coordinated activity. Over to paleontologist and evolutionary biologist Nicholas Longrich:

Yes, let’s consider ants!

Just over a year ago, a team from the Weizmann Institute of Science published a delightful study of individual versus collective intelligence that pitted humans versus ants on a very specific cognitive challenge. The task at hand? A version of what’s called a “piano mover’s” puzzle, having to maneuver an awkward and bulky physical object through a tight space. In this study, both humans and ants had to move a T-shaped beam through a relatively narrow corridor, as pictured above. You’ll get an even better sense of the competition by watching this video:

Both humans and ants were tested in three different “solver” scenarios: (1) individual humans versus individual ants; (2) small groups of humans and ants (six to nine each), and (3) large groups of humans (around 25) versus large groups of ants (around 80). Additionally, for the group scenarios, in some trials humans were allowed to communicate talk with each other, but in others they had to wear sunglasses and masks so as to limit their communicative abilities to pushing and pulling on the beam, as the ants did.

What the researchers found will not surprise anyone who’s ever tried to coordinate a multi-family camping trip. In the first scenario, individual humans beat individual ants across the board—hooray us. But in the group scenarios, however, ant performance improved with greater numbers, whereas human performance decreased. What’s more, when human communication was restricted, ants actually solved the task faster than we did. In other words:

Alert readers may notice there’s a seeming contradiction here. On the one hand, I’m suggesting that our ancient ancestors may have triumphed over Neanderthals because of our greater collective intelligence. Yet, on the other, I’m pointing to evidence that suggests that our capacity to solve certain types of problems actually drops as more of us gather together. What gives?

Well, both of these things can be true, and that’s the point I’m driving at. Under certain circumstances, such as, say, navigating complex European ecosystems undergoing climate change, collective H sapien intelligence may have proven highly advantageous. But under very different conditions, such as, say, trying to move a weirdly shaped object through a cramped space, our individual intelligence works against us. There is no universal truth about intelligence, collective or individual, that suggests more is always better. We need both nuance and theories as to why certain types may be advantageous in particular settings.

Which brings me to my obligatory AI-related commentary. As I’ve argued previously, Big Tech CEOs such as Sam Altman and Dario Amodei explicitly conceive of intelligence as a commodity, a good to be traded on the open market and priced in per-unit costs. But even if we grant for sake of argument—for sake of argument!—that AI represents some novel form of cognitive capability manifesting in digital form, that in no way means that it will prove to be universally beneficial for any particular problem we humans want to solve. Software engineer Simon Willison neatly illustrated this tension when he recently observed that, using AI tools, he can now produced 10x as much code in a day but no longer has the capacity to review it, leaving him feeling both guilty and worried about what might happen if something goes wrong. And we’ve already seen things go wrong.

Zooming out a bit, Henry Farrell and Cosma Rohilla Shalizi recently published an excellent analysis of AI as social technology that touches on this theme. They begin by noting that various human institutions are designed to foster collective intelligence by channeling “the production and gathering of knowledge across many individually limited minds”—the birth of bureaucracy. The downside, however, is that these “vast systems [] are regularly indifferent, and sometimes inimical, to the particular fates and desires of individuals and groups.” Indeed. From this foundation they explore myriad questions of how AI might intersect with various forms of bureacracy management, highlighting the many “known unknowns” before reaching this conclusion:

There is no universal method of solving complex problems through some form of digitized, commodified intelligence. And while humanity benefits from our capacity to share our knowledge with each other, our T-beam competition with ants demonstrates that our intelligence can work against us too. So choices will have to be made. What’s becoming clearer by the day, though, is that voices of collective concern about AI in society are getting louder and louder.

I mean, just listen.

Subscribe now