AI-powered
podcast player
Listen to all your favourite podcasts with AI-powered features
The Problem of Analogy in Machine Learning
M phd. work with douglas hofstetter was on getting computers to make analogies in an idealized domain, you know, sort of micro world. The goal wasn't to get a programme that could make string analogies, but rather to capture something about real world analogies in this idealized domain. Other people have tried other kinds of approaches, but nobody's been able to solve even these simple seeming problems with the same generality that humans can.