Matt's Thoughts In Between - Issue #5

One of the topics in the Holden Karnofsky podcast I mentioned last week that I ran out of room to talk about was Utopias - and particularly the idea that it’s very hard to describe an ideal society that is actually appealing. With this vaguely in the back of my mind, I stumbled upon this excellent episode of John Danaher’s podcast on the topic. It’s an intriguing discussion of the role that games might play in Utopia, based on The Grasshopper by Bernard Suits (which I’ve now ordered but not yet read).

The basic idea is that in a post-scarcity world, where all instrumental problems are solved, games will be all we have left (though if abundance comes from general AI, won’t the AI be better at the games than us too?) It’s a neat idea, but not really an inspiring one. As I’ve mentioned before, I’m very interested in the apparent lack of positive visions of the future in evidence today. If you have a favourite one or have read something interesting on this topic, hit reply and let me know.

It’s hard to talk about the future of AI today without talking about China, as mentioned in previous issues. Despite this, there’s a dearth of information on what’s actually happening in Chinese AI (or “AI with Chinese characteristics”, as someone wittily put it). So I was excited to stumble upon Jeffrey Ding’s Twitter account. Ding does China AI research for the Future of Humanity Institute and has a superb thread translating a recent 500-page Chinese book by Tencent and a government think tank on AI policy. The result is full of nuance and interesting details, a far cry from some of the “arms race”-style reporting on the topic in recent weeks. Ding has just started a newsletter on Chinese AI, which I think will become a must-read.

Related: This game makes you feel like an Reinforcement Learning agent playing a game for the first time by removing all the cues that humans are used to. Both scientifically and anthropologically fascinating.

Slate Star Codex has a (relatively) new essay, Conflict vs Mistake, exploring whether we should think about political disagreement as the result of different views of facts and arguments (“Mistake”) or irreducibly opposed interests (“Conflict”). This got me thinking about where political change comes from. My tentative view is that many major political fissures are best understood as intra-elite conflicts - even (or perhaps especially) when they’re presented as elite vs populist conflicts or pure ideological disagreement. 

What’s fascinating is that elites often cannot contain the forces they unleash as they try to best their opponents. In this view, Brexit is a continent-scale disruption born out of a internal dispute in the British Conservative Party and the Reformation was a series of dynastic struggles in Holy Roman Empire played out as a grand theological dispute. 

Related: If, like me, you’re a sucker for this type of sweeping historical analogy, I recommend this podcast (sent to me by my colleague Freddie fforde) on the links between the Reformation and Facebook…

Thanks for all your “re-reading lists” last week. Please do hit reply if you have recommendations of things to read or any comments. And if you like this, please tell a friend or share the subscribe link: https://www.getrevue.co/profile/mattclifford

Until next week,

Matt