David
twothreemany.bsky.social
David
@twothreemany.bsky.social
Tech worker. Bay area native. Jew-ish. Recovering sectarian. U+03C2 Greek Small Letter Final Sigma male.
Yup!
February 8, 2026 at 11:44 PM
Tesla and Waymo are not the same (though if people are unable to distinguish that does incentive Waymo to give as little of a shit as Musk, an incentive to which I really hope they won't respond)
February 6, 2026 at 7:51 PM
Liking Stalin is edgier than liking Khrushchev so the Twitter tankie position is easy to predict. (On the other hand liking Khrushchev specifically would be nicely kneejerk contrarian; is this in Yglesias' ouvre somewhere?)
February 5, 2026 at 10:48 PM
No question there's a degradation in standards for politicians
February 5, 2026 at 7:09 PM
*median undecided, at least
February 5, 2026 at 2:33 PM
I mean I have no doubt I'm a deeper thinker than the median voter and so are you and my elderly cat at least has a case, but I don't know if that's new
February 5, 2026 at 2:31 PM
Fair. This one is tempting (or I wouldn't have replied) but I don't know how to tell if it explains anything or just feels good because it's nice to feel like I'm smarter than my political opponents and not losing a cultural fight
February 5, 2026 at 2:29 PM
Mm seems unfalsifiable?
February 5, 2026 at 2:23 PM
(61%)
February 5, 2026 at 6:27 AM
I am less TESCREAL than Liam Bright, does this make me negative racist?
February 5, 2026 at 6:26 AM
I don't think it's likely a recession ends up net positive for the party in power, even an authoritarian one. Lotta dictatorships have fallen during economic hard times
February 5, 2026 at 4:28 AM
I don't see Bender here doing "remainder humanism", at least not in the podcast quote? She seems instead to be mad about someone admitting they can't refute solipsism, which is silly, yes, but tendentious moralism of a kind not really related to AI
February 4, 2026 at 5:55 AM
The good ones can often tell you a lot about their mistakes as well as other people's
February 3, 2026 at 7:22 PM
There is something to learn from people who have been in the fight for a long time about what _doesn't_ work, and sometimes about the deep causes of the problem! But less about what _does_ work
February 3, 2026 at 3:30 PM
Double take at "professor of business administration" - I forgot this guy's data fabrication hadn't already sunk his career!
February 3, 2026 at 2:46 PM
I don't think these folks' theory of change involves voters & voting. Not to say their alternative strategic picture is coherent. Last I had some in my social circles, a decade ago, The Coming Insurrection was a typical influence--lots of poetic language as substitute for structured argument
February 3, 2026 at 2:42 PM
DOGE actually briefly looked like it was going to kill the US AI Safety Institute but all it did in the end was force a rebrand
February 1, 2026 at 11:35 AM
I'm someone who believes broadly that AI is very plausibly the most important story in the world right now but I don't really see that case for Moltbook... Is it not just AI Village but less responsible?
January 31, 2026 at 4:25 PM
Reposted by David
For example, we spent 100s of millions of dollars on a field hospital in Gaza (very much not Best Practice) so that Israel could go in front of the ICJ and say that they're not committing genocide because, look, we let USAID fund this field hospital

They bombed the field hospital repeatedly btw
January 30, 2026 at 9:51 PM
Are you capable of recognizing which of your opinions are correct, though? (Hint, it is related to which are also my opinions)
January 31, 2026 at 2:39 AM
I would love for them to have just forgotten how to make good models but I'm not betting on it
January 31, 2026 at 2:12 AM