Seeing Like a School
• ❈ •
A useful perspective to have on schools, over and above their various benefits and deficiencies, is that they are altogether rather strange.
People—teachers, students, parents—have been complaining for a century that memorization is pointless when ”you can just look it up“.
This complaint predates AI, it predates Google, it predates this internet.
But it’s wrong. Here’s Pauling on why he gave closed notes exams:
5yo got a new classmate in her Montessori school this year, a 6yo girl from Korea.
6yo didn’t speak any English at first. But 5yo was quite drawn to her. Just a very different child than what she’s used to—overall calm, precise and skilled with e.g. drawing, coloring, folding.
My most bespoke parenting opinion is that big screens are perfectly fine for kids but small screens are bad. We have a projector in our living room with a huge 6'x10' screen. When the kids watch things on it, they are in motion. They roll around giggling; they climb on the couch,
> Why do your children talk like nerds?
2yo: Fix this door
me: ok *attaches door to toy car, passenger side*
2yo: NO. It goes on the driver’s side
me: If you try that I think you will find that it has the wrong chirality
2yo: Wrong chirality? *tries*
2yo: Oh, ya
This one social relationship was very impactful for my daughter.
Her Montessori classroom is quite good, but I don’t think this year would have been nearly as valuable for her if it hadn‘t been for the 6yo.
Cohort effects can be strong even in cohorts of 2.
I had almost this exact experience with 5yo. She didn’t think I was joking exactly but she was incredulous. “A *hundred* *explosions* every *second*???”
5yo taught 6yo English every day, and reported back to us on 6yo’s progress.
In turn, 5yo was inspired to e.g. “color like a grownup just like [6yo] does”. Really elevated 5yo’s care and her efforts towards dexterity and mastery.
Most white collar jobs aren’t 8 hours of paradigmatic-letter-of-the-job-scope production. There’s lots of collaboration, lots of scope fuzziness, lots of opportunities to do different things, just like there would be in any project. Maybe this is just your point.
I like the idea of there being a million idiosyncratic personal secret benchmarks for AI that no one talks about until they are met. Democratized Goodhart antibodies