Note — Jul 25, 2021

The Computer Scientist Training AI to Think with Analogies

Seen in → No.183

Source → quantamagazine.org/melanie-mitchell-trains-...

At Quanta, an interview with Melanie Mitchell who has worked on ‘digital minds’ for decades and who’s focus is on making AI understand analogies. The technology aspect is worthy of the read but I especially enjoyed the outlook it gives us on our own intelligence, our own understanding.

Think of it in parallel with Webb’s post above, how we use language and how we understand through analogies, and his thinking about the use of words as it relates to this by Mitchell: “‘understand’ is one of these suitcase words that no one agrees what it really means—almost like a placeholder for mental phenomena that we can’t explain yet.

The interview closes with the ideas of embodiment and of combining this ‘analogical’ approach with deep learning. That reminded me of our own modes of thinking (see the power of walking or divergent thinking), and had me wondering whether combining two modes has been done in AI.

Mitchell maintains that analogy can go much deeper than exam-style pattern matching. “It’s understanding the essence of a situation by mapping it to another situation that is already understood,” she said. “If you tell me a story and I say, ‘Oh, the same thing happened to me,’ literally the same thing did not happen to me that happened to you, but I can make a mapping that makes it seem very analogous […]

You can show a deep neural network millions of pictures of bridges, for example, and it can probably recognize a new picture of a bridge over a river or something. But it can never abstract the notion of “bridge” to, say, our concept of bridging the gender gap. These networks, it turns out, don’t learn how to abstract. There’s something missing. And people are only sort of grappling now with that. […]
I think this mechanism of abstraction and analogy is key to what we humans call understanding. It is a mechanism by which understanding occurs. We’re able to take something we already know in some way and map it to something new. […]

One of the most important things for you to do is to model what other people are thinking, understand their goals and predict what they’re going to do. And that’s something you do by analogy to yourself.