10 Comments
User's avatar
Joseph Thibault's avatar

Interesting that the default is "10th grade", which puts it out of cognitive reach for 90% of the students that the union supports (k-12) and likely speaks more to the demographics of the highest use case (college)

Expand full comment
Benjamin Riley's avatar

Great observation. Part of the challenge of building a "do anything for anyone at anytime" product is that it's just about the opposite of what people learning need.

Expand full comment
Christa Albrecht-Crane's avatar

But I also find this 10th grade default infantilizing for college students or adults in general; it's literally dumbing down explanations or summaries, or whatever it's generating. As Benjamin Riley says in his comment, these general catch-all programs end up serving no user well because they cannot work well for generic tasks.

Expand full comment
blindboy's avatar

There has been a fundamental failure by governments and education bureaucrats over many decades to develop digital systems that support teachers and students. Most are just expensive ways of wasting time. My observations: 1. Most high school material is best taught using traditional methods. 2. Schools have been exploited by software companies to condition students to use their products. 3. Student engagement improves 1000% when technology is removed. 4. Schools which ban mobile phones then encourage the use of laptops with virtually unrestricted internet access are kidding themselves.

Expand full comment
Christine Drew's avatar

Do teachers use the science of learning? Would it not be a great research project to observe and annotate teachers' interactions with students vis-à-vis this tool? What if teachers used the tool on a whiteboard with their class and talked about it as they created something together? What could we/they all learn from this? Does the current teacher shortage make this more appealing than it should be? But, most importantly, are we already forgetting that computer labs, once ubiquitous, are now gone...A pilot might be in order, not a rush to adopt everywhere, led by those with a profitable interest. Thanks, Ben, for the scenario that should be a caution to all of us who care about kids and the future.

I'm reminded of the question of ten or so years ago, "If I can google it, why learn it?" Now, it's "If I can get AI to create it, why do it?" We seem to forget, again, that the basis for creativity and innovation is a deep and broad platform of knowledge, resident in our brains. Creative people reassemble or combine old knowledge into new things. Creating curiosity to learn might be our number one priority right now. I think I'll take my grandkids fishing and see what they can learn from experiencing and observing. Then we can talk about how that compares to using AI....

Expand full comment
Benjamin Riley's avatar

I wish we were taking the research-based approach to AI in education that you outline here, Christine! But we're doing the opposite of that -- look how proudly Belsky boasts about the student use of ChatGPT. It amazes me that they are excited to harm the learning of so many kids.

Expand full comment
_ikaruga_'s avatar

Their excitement over the harming "of so many" people's (why kids only? So-called adults are no less lazy, or naturally inimical to their own learning and understanding) learning pales next to that of the average pedagogue/"education scientist" from the last 100 years, and unlkme those vain virtue signallers, the tech people at OpenAI and elsewhere don't live feeding themselves and their "followers" denials of truth regarding human nature: which means they are going to make forecasts on behaviour, consequences, outcomes, that are in a relation with reality, unlike the pedagogues', and the Deweys of this world's.

School as it once did has stopped existing, and its "democraticity" will ensure it will keep not existing.

The few who are able and willing to study and learn are blessed to live in the time o

where Internet, then also AI, entered the history of our civilisation.

Expand full comment
Christine Drew's avatar

The best answers to our questions lie not in confirmation bias or binary thinking (either/or decisions), but in slowing down our responses to consider many and diverse points of view. Better outcomes for our students could come from conversations and collaborations between 'pedagogues' technologists, and data scientists. It's time educators, technologists, and researchers work together for the good of society. There's an art and a science to teaching and learning. It's a craft and a practice, honed over time. Maybe there's an oath for us all: But, first and foremost: Do No Harm.

Expand full comment
Will Granger's avatar

Anyone who thinks OpenAI knows anything about education or that it cares about students is naiive or complicit in caring about profits over student well being

Expand full comment
Sepehr Vakil's avatar

Beyond metacognition and inquiry base learning, the learning sciences has made major advances in understanding learning as fundamentally tied to culture, identity, ethics, and emotions. Current AI system do not even pretend to understand these major breakthroughs and are therefore miserably disconnected from cutting edge scientific research on Learning

Expand full comment