Towards a Universal Schelling Machine
In which Tom Schelling and I compare notes on the strategic implications of AI
NotebookLM is a powerful AI tool from Google. I tasked it with doing some fairly tricky analysis of strategic theory. And it came up trumps - you can listen to its thoughts below. But first, read on:
By now, I’ve written a fair bit about strategy - lots on its psychological essence, and some on the impact that AI might have on that. Tom Schelling knew a thing or two about the psychology of strategy and also thought deeply about the impact of revolutionary new technologies - in his case nuclear weapons. (Sidebar - I don’t think these are all that revolutionary).
I wanted to see how far our thinking was complementary, or otherwise. Were there salient themes that we might learn from thinking psychologically about the strategic implications of AI, along the lines suggested by Schelling and me? Turns out there were.
That’s interesting in itself, but just as interesting to me is the app that unearthed them.
Notebook lets you upload sources - I used book chapters, articles, videos of our lectures. I could have used audio too - next time. In its experimental mode, the app lets you upload up to 50 of these. Then you can interrogate it, via a chat window - I had much fun, and learned a good deal. Easily enough for a new article, I reckon.
Of course, you can also generate one of those AI podcasts that have been getting lots of online attention. They are superb. I’d certainly be happy with my students listening to the one it made about AI, Schelling, and me. You can hear it below - see what you reckon.
The whole thing raises important questions for me, as a scholar, about what it is we’re doing here - and also as a teacher, about how and why students should be learning, and what they should produce. Here’s that podcast.