Astral Codex Ten Podcast

Since the beginning of time, mankind has yearned to drink a cup of coffee at 5 PM and go to sleep at 10. But now there's an entire scientific subfield and several companies pandering to this absurd fantasy.
Here's a simplified diagram of caffeine metabolism:
Caffeine is converted to a different stimulating chemical called paraxanthine by a liver enzyme called CYP1A2, over a half-life of ~5 hours. Then the same enzyme converts paraxanthine into other non-stimulating chemicals that don't matter, over a half-life of about three hours. Kman on Less Wrong has graphed the total effective concentration over time:
Over the course of hours, the liver converts caffeine into paraxanthine. These are (by assumption) equally potent, so total effective concentration (the dashed red line) goes down slowly.
(Why does it go down at all? Because of the simplification in the diagram above: only about 80% of caffeine is converted to paraxanthine; the rest is converted to other, less active metabolites.)
By hour ten, paraxanthine predominates, the liver is mostly converting paraxanthine to other non-stimulating chemicals, and effective concentration continues to decrease. In this model, we see that the effective half-life of caffeine is close to ten hours1!
What if we want it to be less? There are three levers we can pull: the enzyme, the metabolite, and the starting chemical.
https://www.astralcodexten.com/p/the-quest-for-caffeine-you-can-have

[previously in series: 1, 2, 3, 4, 5, 6, 7, 8, 9]
Paul Graham once said that every city sends a message. New York says you should make more money. Berkeley says you should live better. Boston says you should be smarter. People shouldn't decide on a permanent residence until they've lived in several cities and understand their differing psychological effects.
San Francisco sends a message too. It speaks it in a cacophony of dissonant voices, some sounding like the hissing of snakes, others like the chittering of insects. "You should pierce the veil," it says, with a faint accent which you are not scholarly enough to recognize as ancient Sumerian. "You should rip through the flimsy screen that separates your world from the infinite, and witness what lies behind. Because you'd like what you saw there? Oh no, nothing like that. But aren't you curious? The thirst for knowledge that moved Eve, Faust, Pandora - don't you have it too?"
There's also a second message from a second voice, one which sounds like a deranged carnival barker, constantly shouting "STONKS! STONKS! STONKS! STONKS!" Which of the two voices you hear depends on your personality, and maybe how many psychedelics you took in college. According to legend, Sam Altman hears both voices simultaneously all the time, like those Tibetan throat singers who can harmonize with themselves. But when the demonic babble becomes too much to tolerate, the San Franciscans drown it out with alcohol, loud music, and various forms of tedious socialization. Thus the famous Bay Area house parties, one of which you will be attending this very night.
https://www.astralcodexten.com/p/the-foothills-of-bay-area-house-party

Old Scottish joke:
A man at the bar is complaining: "You see that bridge there? I built it with my own hands! But do they call me 'MacGregor The Bridge-Builder'? No! And I raised five beautiful children! But do they call me 'MacGregor The Child-Raiser'? No! But, you fuck one sheep . . ."
Has MacGregor been treated unfairly?
https://www.astralcodexten.com/p/macgregor-the-bridge-builder

Last month, some of Silicon Valley's biggest companies signed an open letter supporting open-weights AI.
Open weights AI is like open-source software, where the creator makes the raw code publicly available for free download. It's good insofar as it's the only way an AI can truly be the user's property, as opposed to something that companies like OpenAI or Anthropic temporarily let you use subject to their corporate guidelines and increasingly-nanny-state-like restrictions. If AI becomes the linchpin of the future, open weights AI feels like the sort of thing that could be the difference between being free yeomen vs. corporate serfs.
It's bad insofar as it removes the possibility of gatekeeping and lets criminals commit crimes with it. Open weights AI could be used for hacking, child pornography, harassment, or terrorism (the weights can't commit the terrorism themselves, but they could give bomb-making or bioweapon-making advice). Since AIs have gotten very good - maybe superhuman - at hacking lately, the specter of a world where anyone can hack any site has gotten people grumbling that maybe open weights should be banned. It doesn't help that China produces the best open weights AI, making the idea seem foreign and almost unpatriotic. Proponents counter that "when AI is outlawed, only outlaws will have AI", arguing that bad people will get open weights AI regardless, and good people can use open weights AI to defend themselves. With the recent open letter, companies including Microsoft, NVIDIA, OpenAI, Intel, Amazon, Meta, Hugging Face, and over a hundred others have come out in favor of this position.
Who's leading the other side? Nobody's admitted to it. Some parts of the Trump administration lean anti-open-weights on China hawk grounds, but have stopped short of explicitly asking for a full ban. Anthropic, the most notable omission on the pro-open-weights letter, made an ambiguous statement supporting "open-weights models that don't have dangerous capabilities" - but the industry expects open weights models to have dangerous hacking capabilities within a year, and AFAICT the letter didn't address that beyond inviting readers to draw the obvious conclusion.
In the absence of a more obvious opponent, some open weights supporters suspect our conspiracy - the loose band of AI safety advocates, effective altruists, rationalists, and pause activists who worry about existential risk from superintelligence. This is a reasonable inference. By design, open weights AI is outside centralized control, and so impossible to permanently align against either human misuse (eg terrorism) or loss of control (eg AI turning against humans). Even if its creator trains it not to hack, anybody in the world can download the weights and retrain the AI to hack all day long.
But in fact, most AI safety organizations have remained quietly neutral, and I don't know of any who make this a centerpiece of their activism1 (though I'm not 100% up-to-date on the whole landscape; if you know of one, tell me). A few have proposed policies that are contingently incompatible with open weights AI existing, but they all frame it as collateral damage rather than something they're excited about eliminating.
I'm also neutral about open weights AI. I think it probably won't be long-term sustainable, but I'm happy to wait for this to become clear in the normal course of things rather than expend effort and political capital to ban it immediately.
https://www.astralcodexten.com/p/open-questions-on-open-weights

Commenters convinced me I made a subtle mis-step in my framing of Psychology Research Is Mostly Fine.
My framing emphasized that social priming studies had failed to replicate, but that many other subfields of psychology hadn't. This naturally led commenters to point to failed replications in other subfields. Although none of these were as bad as social priming, few or none were entirely problem-free.
Instead of emphasizing the difference between social priming and other subfields, I should have emphasized the difference between settled science and novel research.
Here are some of my favorite psychology findings:
https://www.astralcodexten.com/p/the-beauty-of-settled-science