Reminds me of articles worrying that AI might find the tasks we give them tedious when... you know... millions of human workers are working insufferable conditions.

Do you see any meaningful differences though?

AI didn't break education or make students cheat. It highlighted the changes we should have made decades ago. So at the very least, can AI be the incentive to governments and educators to rethink foundations that cannot be solved with virtual tutors that are hidden behind two drop-down? Please?

While waiting on policy to change is likely to take much longer than I'd like, I'm still surprised and the reactions from academic institutions. From simple reactions to either completely ban AI or adopt it blindly, to simply doing nothing while taking years to revise curriculum.

The alternative of turning to companies and hoping their teams fix things is not of interest to me. I don't think offloading the responsibility of selecting learning modes, and choosing to ignore grades in favour of more cognitively rewarding learning methods should fall on the students.

But my prototype is not a revolution. It's barely a minor step. And I'm actually quite happy with that. Multiple players need to make a lot of minor steps to make change within massive systems like education.

This is not to say that technologies like AI can't bring in some improvements to education. Quite the contrary, technology in general COULD play a part in facilitating the outcomes we want from teaching and learning, and I'm actively exploring ideas like this here: youtu.be/5-0OolTWm8k?...

It's the misaligned incentives in academia. So long as we keep grading and assessing students as we currently are, education will be a game of simple metrics. Meaning, you can expect students to prioritize behaviours that might reduce learning (unintentionally) in favour of improving their grades.

If you're a designer, I bet that your answer is:

"Design the interface so that it's easier to find. We need to reduce the friction towards good behaviour."

To which I would say you're half right! We SHOULD reduce the friction to good behaviour! But the friction is not ChatGPT's design.

This is true! Though like with a lot of science the packaging and delivery matters quite a bit.

Have you come across techniques or cases where change happened at an institution or with individuals after learning about certain cogsci discoveries?

You're closing the loop. This might become a legitimate new mental model for me 😂.

The only real difference is unlike the group, it never blames the lack of laughs on bad prompts 😬

My own journey of becoming an instructor and facilitator through intuition supports this! As I learned about cognitive science, I find that the result is "obvious" but the formalization is incredibly powerful. Yet when I convey this to peers, they shrug it off like "yeah, so what…"

Brings to mind the question “What are the limits of mimicry?” From @beaudoin.social’s latest exploration: bsky.app/profile/beau...

I am not settled on my level of optimism, but I think “parrots” in this context could be just enough to have terrible consequences.

100% one or the best videos I’ve seen tackle such a subject. Clear, objective, and entertaining as heck!

Heh! I had a similar reflection watching a Japanese Taiko drum show yesterday. Drums are probably the first ever instrument we played as humans, and still 10,000 years later, we aggregate in mass to watch people hit them. It’s kind of incredible.

Love this. Unfortunately in some cases it's almost an instinct. Like jumping into the YouTube comments right after a video is over, and having your opinion impacted almost immediately.

I was actually just thinking about something similar… It's very sad to see what people are willing to attack just for their own gain.