2026-01-08
i want to remember it all
Conversations with Tyler interview with Brendan Foody on AI as unlocking aggregation at scale for basically everything, from hiring, to dating, to skills1. This matches my intuitions that everything most people dislike about modern technology and society isn’t really because scale is bad, but more that our current abilities are not good enough yet. Related, Molly Cantillon (via Marginal Revolution) on organizing your life’s information via Claude Code2, and Ashe on using AI as a personal assistant (Edit: also Dean Ball on his use of AI agents).
Beren Millidge on what is necessary for the future to be dominated by singular superintelligence who maintains a decisive strategic advantage over all other intelligences.
Lilly Ottinger in ChinaTalk on Goertek’s position in the AR/VR headset supply chain. Given indications of increasing demand, it will be interesting to see if Meta’s much derided investment in AR finally ends up paying off with rising demand, or if despite paying the significant costs of being early in hardware, they’ve failed to develop any moat within the supply-chain.
Ken Opalo predictions for Africa in the upcoming year in governance and economics. On the intersection of the two, I’m curious if increasing demand for mineral resources will tempt any governments to play the nationalization card. Tangentially related, Bryan Caplan reflections on his debate with Matt Bruenig on private property.
Justin Ross on just doing things instead of worrying about the meaning of your actions3. This is framed as a sort of no-nonsense social conservative perspective, but it works just as well in the social liberal sense of just doing what it is you personally want, as a participant rather than a would-be director of the market of what is.
Isabel (via Sarah Constantine) on the mechanics behind effective manifestation.
No Magic Pill impressions from his Taiwan trip.
Alex Araki linkthread4.
Eleanor Konik linkthread.
Also, Hollis Robbins comments on the opening question of this interview on hiring poets to improve LLM poetry skills, and to what extent the ability to reliably produce quality outputs can also produce greatness.
Edit: Some people are not a fan of this: it’s not clear to me if it’s because they view it as untruthful hype, or if it represents a future that they do not want to see. I don’t use Claude code to organize my life (yet), but I see this as sort of parallel to Bryan Johnson, in that I might in the future. In which case, biased as it is, it seems to me that it’s still a useful datapoint.
Unless you personally enjoy online discourse as I do, in which case you should do it with full awareness that you’re primarily doing it for fun and not let it pollute your real life interactions. On that note, Sam Senchal has a post on his understandign of Observer Theory, combining the Ruliad with the concept of ideological egregores. Possibly coincidentally, Andrés Gómez-Emilsson argues that Ruliad-based views of reality do not sufficiently explain consciousness. I personally have no particular view on this topic. There was an interview between Sean Carroll and Ned Block recently which provided a good basic overview on theories of consciousness, and basically all descriptions seem equally unpersuasive to me. For example, the lookup table argument against functionalism seems to break down the instant you start considering some harder inputs, like “what is the next number?” If one tries to model a lookup table which would actually end up satisfying the Turing Test, the resulting tables would probably have become sufficiently complicated, capable of self-reflection and understanding context, that it’s not entirely clear to me that one could dismiss its consciousness out of hand.
Edit: similar argument mentioned by Kore in his post on LLM consciousness.
Which includes an interesting proposal by Denisa Lepadatu and Ed Boyden on a standard model for biology. Given how complex this would be, I’m curious as to what extent this is something that is needed in order to unlock functional immortality, as opposed to being produced as a side effect and then fully instantiated after the fact, as a sort of crowning jewel.