Think outloud for https://nitter.cf/t.co/1Dbt4zwdm2 https://nitter.cf/t.co/0pFwy0400C
Joined September 2012
- Tweets47.6K
- Following3.2K
- Followers46.7K
- Likes56.6K
Eric Levitz retweeted
I'm with her
Hillary Clinton: “His surrogates and a lot of the independent expenditure media is really fear-mongering against El-Sayed as a Muslim, trying to link him to the Muslim Brotherhood, Trump is doing that. I’m just tired of that and I hope people are tired of that. Judge somebody on who they are, what they stand for, what they produce. This is not about woke, this is about broke. If you can’t win on affordability, if you are bankrupting the country and enriching yourself personally, try to divert attention”
As with a lot of David's stuff, I think the conceptual framework here is brilliant but his case for it is undermined by wild overstatement + dubious examples.
The argument that people don't earnestly believe, deep down, that education confers economically useful skills (rather than merely credentials with signaling value) seems very weak on a variety of levels.
For one thing, formal education obviously does confer such skills; it is not coincidence that the spread of public education coincided with gains in literacy. And although some of pre-med/pre-law education consists of filtering mechanisms to maintain labor scarcity, it seems clear that doctors and lawyers rely on skills acquired in school to practice their trade.
For another, the Caplan argument is very silly. It implies:
1) That people sincerely believe that a somewhat tedious and effortful activity would be good for them in the long run, they'll never have a positive emotional reaction to gaining reprieve from that activity in the short term. It's like saying that, if someone feels happy when their dentist appointment is cancelled and their afternoon is suddenly freed up, they must not really believe that dental cleanings are good for their teeth.
2) That students' reacting happily to a single class session being cancelled suggests that they would be similarly enthused if every session of all their classes were cancelled -- such that their tuition bought them no actual pedagogy -- so long as they still received a diploma. This does not seem like a sound inference!
Yeah. In general, I think people are just really used to arguing about much larger demographic groups (Black voters, Latinos, the white working-class etc). So when there's competition for other ethnic blocs' loyalties, people default to the same modes of discourse (also, I think Jews' overrepresentation in media/politics gives many an exaggerated sense of how numerous we are in the broader country)
Jews comprise roughly 1.4% of the Michigan electorate. Muslims account for about 2.4%. And only a fraction of either group are swing voters.
Every vote matters. But The Discourse exaggerates the importance of Michigan Jews in 2026 (and of the state's Muslim voters in 2024)
.@VanJones68 says Abdul El-Syed has a "real problem" winning over black and Jewish voters in Michigan.
"Part of it is self inflicted...he needs to denounce these ads and put this coalition back together."
@ScottJenningsKY: "the reason he gets attacked like this...is he brings it on himself."
More via @CNNTheStoryIs: youtube.com/watch?v=RPpgI_jI…
I don't think it's true that Claude was ever producing prose that was *too* clever and concise for mere mortals. As a writer, it is superhuman in its speed, range, and perhaps accessibility. But stylistically, all the LLMs have always been "off" in ways both blatant and indescribable.
That shortcoming arguably doesn't matter that much though. The robots' strengths still enable them to outcompete human writers for most commercial purposes.
What would the writing of superintelligent AI sound like? You might expect it to have the characteristics of good human writing — figurative language, high information density, coining new terms, and punchiness — but dialed up to 11. Well, that’s what Claude did, until Anthropic dialed it back recently because of user complaints. It didn’t sound superintelligent to most of us. Instead, the neuralese writing style sounds like a “stupid person’s idea of a smart person”.
Our response to information density and metaphor-laden writing looks something like the cartoon graph I've drawn. Beyond a point, it completely backfires. The reason human writers don’t overdo it isn’t because they aren’t capable, but because they are aware of the reader’s cognitive burden. One of the hardest skills to learn as a writer is to simulate the reader’s experience.
It’s easy to train AI to mimic mediocre human writing. But optimizing it to be (perceived as) good is really hard, and made much harder by the fact that when everyone uses the same few AI models we get tired of their tics very quickly. Most Claudeisms are individually pretty good. I hate that I can’t use “load bearing” in my writing any more. If it weren’t for AI overuse it would be a neat metaphor.
A broader question is whether most tasks in the economy are like chess, where there is a clear performance metric that can be optimized, or whether they’re like writing, where quality is a multidimensional balancing act without clear training signals beyond human mimicry.
Those alleging that El-Sayed supports the Muslim Brotherhood often point to a 2012 editorial he wrote, calling on America to support Egypt's democratic revolution.
It is telling that this is the best they can do. In that piece, El-Sayed makes clear that his sympathies lie with democracy, not Islamist movements, which he describes as a "symptom" of dysfunctional government; people generally do not describe their own ideologies as regrettable markers of a disease huffpost.com/entry/why-the-u…
Eric Levitz retweeted
As @EricLevitz wrote earlier, it would take a LOT to stop the data center buildout vox.com/politics/501146/data…
Wow.
Business investment surged almost 13% in Q3 2026 due to the AI investment boom.
We haven't seen anything that strong since Q3 2011 when the economy was finally starting to pickup a bit of momentum after the Great Recession.
This will not be sustainable at this pace.
[Note: I'm looking at "non-residential fixed investment" and I'm excluding the 2020 pandemic re-opening]
Shouldn't need to be said, but there is 0 evidence that El-Sayed sympathizes with the Muslim Brotherhood, much less that he is some kind of sleeper agent. He is a progressive Democrat who supports abortion and LGBT rights, not a reactionary theocrat (unlike some Texas senators)
Sen. @tedcruz (R-TX) says that if U.S. Sen. candidate Abdul El-Sayed (D-MI) is elected into the Senate and is placed in the Senate Intelligence Committee, "there is a very real risk that Americans will die as a result."
I take him to be arguing that rich white people have outsize influence over what is considered prestigious and will downgrade the prestige of the Ivies in response to this development out of their (perhaps not entirely conscious) racism, rather than saying that he personally considers the Ivies less worthy of prestige because they've become more Asian.
Eric Levitz retweeted
Flat out ethno-racist hysteria and demagoguery.
Sit this one out and this is whose side you’re on.
Sen. @tedcruz (R-TX) says that if U.S. Sen. candidate Abdul El-Sayed (D-MI) is elected into the Senate and is placed in the Senate Intelligence Committee, "there is a very real risk that Americans will die as a result."
Eric Levitz retweeted
There's a leftist/academic thing to constantly emphasize both the moral brutality of capitalist growth in the US and the surprising growth under brutal communism in China.
I think it's good to see the full picture of reality, but the emphasis is …. notable. Leftists/academics are never tweeting about how “did you know America’s economic growth was surprisingly robust when child labor was legal?”
Under “dictator” Mao, life expectancy in China increased by 30 years, literacy rates went from 20% to 70%, steel output grew from 1 million to 23 million tons, 47 million hectares of land was taken from landlords and given to peasants, and China was unified under a strong central state and got nuclear weapons.
Despite his failures, it does not take a rocket scientist to understand why Mao is revered in China.
Readers added context they thought people might want to know
Mao's Great Leap Forward and related policies caused an estimated 30-45 million famine deaths in 1959-1961 alone, with total excess deaths under his rule often estimated at 40 million or higher.
en.wikipedia.org/wiki/Great_Chi…
pmc.ncbi.nlm.nih.gov/articles/PMC43…
chinafile.com/library/nyrb-c…
I initially typed "(a.k.a. going 'full Synecdoche')" at the end of that tweet, but worried it be annoying. Probably my favorite movie
Funny to think about how well-funded Nathan Fielder’s next project is going to be. Dude might actually build an entire model city
Hate electing to kick in overtime. The knowledge of what you need isn’t worth giving the other team the potential third possession imo
Eric Levitz retweeted
The world has incredible detail, friction, and complexity. People good at understanding machine learning and predicting benchmark performance of AI models don't necessarily have a good understanding of these factors and instead often couple impressive forecasts of AI capabilities with low-resolution folk models of society.
The result is many silly forecasts of AI's allegedly explosive societal impacts when it comes to economics, the information environment, etc, among AI experts and insiders. Good thread 👇
