The AI Superforecasters Are Here
Astral Codex Ten Podcast
The annual was earlier this month. This was the year prediction markets went from an obscure hobby to a multi-billion dollar industry; from semi-illegal to having the President’s son as an advisor. I can’t remember if anyone talked about any of that. It didn’t even register. All eyes were on the AI superforecasters. I met an AI superforecaster startup founder who told me his AI had turned $35 into $2 million on Kalshi over seven months. I met another who said they were beating the stock market by 25% with a market-neutral portfolio - of course this could be luck, but they’d beaten...
info_outline
Chip Off The Old Block
Astral Codex Ten Podcast
I. Having kids has given me new appreciation for old poetry. The first time I read Song of Hiawatha, I skimmed over the part in Book 3 where Hiawatha first meets his father Mudjekeewis: Filled with joy was Mudjekeewis When he looked on Hiawatha, Saw his youth rise up before him In the face of Hiawatha, Saw the beauty of Wenonah From the grave rise up before him. "Welcome!" said he, "Hiawatha, To the kingdom of the West-Wind! Long have I been waiting for you! Youth is lovely, age is lonely, Youth is fiery, age is frosty; You bring back the days departed, You bring back my youth of passion." ...
info_outline
The Metaculus Threat To Democracy Index
Astral Codex Ten Podcast
In recent posts on and , people have asked me - how do you know you’re not suffering from Trump Derangement Syndrome? I take this seriously; we’ve all lost loved ones to this condition. The best check on my reasoning would be an objective measure of the health of American democracy. There are several “democracy indices” that purport to do this, but they have a mixed reputation. My impression is that most current accusations of bias are relatively weak - I agree with - but they rely enough on “expert” opinion that I don’t expect them to convince a skeptic. The newest entrant in...
info_outline
Should People Avoid Whole-Body Screening Info?
Astral Codex Ten Podcast
The most controversial part of last week’s article on the Midjourney ultrasound scanner was medical experts’ recommendation against whole-body screening (including existing whole-body screening technology using MRI). Isn’t this crazy? Whole-body screening can save lives by detecting serious diseases like cancer. The experts counterargue that it finds so many false positives - minor zit-like imperfections that would never have caused problems, but which cost patients time, money, anxiety, and side effect burden to investigate - that it ends up net negative. But isn’t this just a problem...
info_outline
Preliminary Thoughts On The Midjourney Scanner
Astral Codex Ten Podcast
like that, except from a medium-sized startup instead of a tech giant. Earlier today, they announced a pivot to medical scanners. The new , which they describe as “a bold new kind of machine to reimagine the foundations of healthcare and our relationships to our bodies”, will be a tank of water surrounded by a ring of ultrasound scanners. The patient goes into the tank, the scanners emit ultrasound from all angles, and then some fancy AI reconstructs the echoes into a 3D picture of the body. The result is ultrasound tomography: the same sort of rich data as a CT or MRI, but done via...
info_outline
Waiting For The Miracle
Astral Codex Ten Podcast
In 1917, three children in Fatima, Portugal claimed to have seen the Virgin Mary. They promised she would perform a miracle on a certain day in October. Nearly 100,000 pilgrims arrived, hoping to see whatever happened, and nearly all report that the sun turned pale, changed color, and spun around. Many other writers have investigated the children and their visions, but I was fixated on this sun miracle. Despite popular discussion of “mass hallucinations”, this is AFAICT the only example of tens thousands of people all saying they witnessed the same impossible thing, at the same time. I got...
info_outline
Never Cross a River Four Feet Deep on Average
Astral Codex Ten Podcast
Guest post by Alexander "Sasha" Putilin [This is a guest post by Sasha Putilin. I encourage any ACX grantees who are interested to write about their projects. - SA] The results of my ACX Grants 2024 project are in. The project attempted to replicate the 2023 study . It claimed that if you read a person’s brain waves, figured out an individual peak alpha frequency, and flashed a bright white light at that frequency, then they learned a certain perceptual task faster. Why bother? The result hinted that learning may depend in part on how well the brain keeps its rhythms coordinated. In...
info_outline
My AI Opinions
Astral Codex Ten Podcast
I recently had a minor spat over someone misinterpreting my AI beliefs (see section marked “Update” at the bottom ), so I thought I would list them in one place, so I can refer people when they ask. Timelines Define AGI as AI intelligent enough to do 90% of knowledge work jobs. I think there’s a 25% chance of AGI by 2027, a 50% chance by 2034, and a 75% chance by 2045. Basic argument: In a certain sense, AI is already “smart” enough for this (eg it can answer quantum physics problems, which require higher IQ than most knowledge work). Its remaining limitations are that it’s...
info_outline
Book Review: The Dialectical Imagination
Astral Codex Ten Podcast
The philosophers of the Frankfurt School practiced a technique called negative dialectics, where concepts are defined as much by what you can’t say about them as what you can. Appropriately, the Frankfurt School has ended up defined by what you can’t say about them. You can’t say that they invented a new form of left-wing thought called Cultural Marxism. This would be (according to Wikipedia) , a “far right anti-Semitic conspiracy theory that misinterprets Western Marxism, especially the Frankfurt School, as being responsible for modern progressive movements, identity politics, and...
info_outline
Use AI This Election
Astral Codex Ten Podcast
I’m not saying AI is superintelligent or can decide better than you can. I’m saying that if you - like me - spend an hour or so doing research before voting on local seats, AI can aid that research very effectively. And if you don’t do that research - because you weren’t willing to waste an hour on it before - AI makes it so much faster that you might want to start. I gave Claude a prompt something like (edited for coherence): I’ll be voting in the June 2026 California primary. I’m a centrist liberal abundance YIMBY whose favorite political writers are Kelsey Piper, Matt Yglesias,...
info_outlineI recently had a minor spat over someone misinterpreting my AI beliefs (see section marked “Update” at the bottom here), so I thought I would list them in one place, so I can refer people when they ask.
Timelines1
Define AGI as AI intelligent enough to do 90% of knowledge work jobs. I think there’s a 25% chance of AGI by 20272, a 50% chance by 2034, and a 75% chance by 2045.
Basic argument: In a certain sense, AI is already “smart” enough for this (eg it can answer quantum physics problems, which require higher IQ than most knowledge work). Its remaining limitations are that it’s confused, unagentic, lacks situational awareness, and tends to hallucinate. The METR time horizon graph, and several other related benchmarks/experiments/intuition pumps, suggest it’s improving on time horizons at an (exponential) rate that lets it cross human-level performance sometime around the early end of the schedule above, and subjectively it feels like harder-to-measure constructs like situational awareness are improving about as fast.
Arguments for earlier: recursive self-improvement causes a speedup compared to the trend. This is one of the biggest blank spots in my model: I don’t know how fast RSI will progress, and I don’t think anyone else does either. There’s some function mapping a combination of AI talent and compute to progress, and we don’t know how it behaves in the domain when there’s far more talent than compute available. It could fizzle out completely for lack of compute, or it could go vertical. The AI Futures Project has done some of the best work trying to model this, but even they have low confidence.