Alison Gopnik on Childhood Learning, AI as a Cultural Technology, and Rethinking Nature vs. Nurture

“There's a beautiful idea from computer science and physics called simulated annealing. You have some problem to solve, some space of solutions. One thing you can do is make little changes to what you already know, a low-temperature search. The other kind is a high-temperature search: bounce around the space, try wild, crazy things, a random walk. The strategy in computer science is start with wild, crazy, out-of-the-box high-temperature search, then cool off and fill in the details.”— Alison Gopnik

Creativity & Innovation · Psychology & Behavior · Technology & Engineering

DUR_ENDURING

High temp search then cool: kids do this naturally

“Eric Turkheimer found that you had much more convergence in twin studies in upper-class contexts than in poorer families. The explanation: if you're in a poor family, small differences in your environment can make a big difference in how the rest of your life goes. And that's less likely to be true in a richer family. The effect of having a protective caregiver is that it allows more variability.”— Alison Gopnik

Psychology & Behavior · Economics & Markets · Biology, Ecology & Systems

DUR_ENDURING

Safe environments allow more variability, not less

“Going back to Plato and Aristotle, they talked about the problem of how we can know as much as we do about the world given such a small amount of data. Two solutions: one is that structure couldn't have been learned from the data, it must have been there innately (Plato's past world, Chomsky's evolution). The other solution: it looks like there's abstract structure, but really it's just statistical combinations of the data (like current deep learning AI).”— Alison Gopnik

Philosophy & Reasoning · Psychology & Behavior

DUR_ENDURING

Rational constructivism: neither innate nor purely statistical

“Kids are better Bayesians than scientists. If you have a very peaked prior, a lot of experience giving you reason to believe an assumption is right, then it's rational not to change it when you have just a little bit of evidence. Kids are better at solving problems with unusual outcomes than scientists are.”— Alison Gopnik

Psychology & Behavior · Philosophy & Reasoning · Strategy & Decision Making

DUR_ENDURING

Kids update faster: flatter priors enable learning

“When people say there isn't an effect of the family environment on development, what they mean is they don't see high correlations between kids in the same family and some measure of what an adult is like. Siblings, the non-shared environment, things that don't seem to be the same, play a larger role than the shared environment. If you have two kids with the same parents, you might think the kids are going to be more similar. But if caregiving allows variability, you might expect the opposite.”— Alison Gopnik

Psychology & Behavior

DUR_ENDURING

Good parenting increases divergence, not convergence

“Imagine if we tried to teach baseball the way we teach science. We would tell everybody about great baseball games when they were little. In high school they could throw the ball a lot to second base. In college they could reproduce great baseball plays. But they wouldn't actually get to play the game until they were in graduate school. If you taught baseball that way, you wouldn't think people would be as good at baseball. What kids end up being incredibly good at is going to school.”— Alison Gopnik

Leadership & Management · Creativity & Innovation

DUR_ENDURING

We teach about science, not how to do science

“Consciousness is very unlikely to be a single thing. It's like life: in the 19th century we thought life would turn out to be a simple thing, and there were research programs about what makes a living thing living. It turned out that was just the wrong question. There are many different kinds of processes involved in different things we think of as life. The same is likely true with consciousness.”— Alison Gopnik

Philosophy & Reasoning · Psychology & Behavior

DUR_ENDURING

Consciousness isn't one thing, like life isn't one thing

“My family: there were 6 of us in 11 years. My parents were graduate students. We had this combination of a great deal of warmth, a great deal of love, an enormous amount of stuff around us, books and ideas. We got taken to the Guggenheim when Adam was 3 and I was 4. We were also completely free. We were in regular public schools. We came home after school and did whatever we wanted. The kids were taking care of each other a lot of the time. The result is you get a lot of variation.”— Alison Gopnik

Psychology & Behavior

DUR_ENDURING

Nurture's effect: variance not mean

“My view about generative AI: our intuitive lay conception of how AI works is really misguided. We have this Gollum view: here's this non-living thing we've given a mind to, and that always works out badly. It's going to be superintelligence for good or ill. That's not the right narrative. The right narrative is to think of it as a cultural technology. It's a way of getting information from other people.”— Alison Gopnik

Technology & Engineering · Philosophy & Reasoning

DUR_CONTEXTUAL

AI as cultural tech, not intelligence

“Babies' brains are very plastic. They're surrounded by novelty. They're not just focusing on one thing at a time the way we do when we're grownups. If you think about contexts as grownups where we are in that state, like traveling to a new place or trying to do something new, it's not like we're unconscious. On the contrary, it feels like we're full of experience. We're vividly experiencing the world around us. That's what it's like with babies.”— Alison Gopnik

Psychology & Behavior · Philosophy & Reasoning

DUR_ENDURING

Less memory compression means more vivid present awareness

“Philosophers of science and computer scientists have found systematic ways to talk about how we get from data to theory. Scientists, mostly not consciously but as part of what they do, and little kids, are looking at data and systematically figuring out what kind of structure out there in the world could have caused this pattern of data.”— Alison Gopnik

Philosophy & Reasoning · Psychology & Behavior · Technology & Engineering

DUR_ENDURING

Both kids and scientists infer world structure from data

“People with aphantasia don't retain images very well, but they vividly experience the visual world. They just don't use that experience when trying to imagine something or generate an image. What aphantasia shows is that generating an image when you're solving a problem is sort of epiphenomenal. You're not solving the problem by looking in your head and seeing the picture. My friend Ed Catmull, co-founder of Pixar, has aphantasia. He's an animator who doesn't see pictures inside his head.”— Alison Gopnik

Psychology & Behavior · Creativity & Innovation

DUR_ENDURING

Mental imagery is epiphenomenal, not causal