trey

26.8K posts

trey banner
trey

trey

@Comparativist

#FloridaMan, lifelong educator, lapsed scholar (HKU PhD). 20 years in the Pearl River Delta. writing about systems, AI, infrastructure, history, and theory

Canton Katılım Aralık 2008
7.2K Takip Edilen8.3K Takipçiler
Sabitlenmiş Tweet
trey
trey@Comparativist·
I hate that China discourse is either Evil Empire 2.0 or Panda Hugger but tbh the discursive space for ppl living under its rule who could provide nuance only allows for Panda Hugger or [bad things happen to you] if you have anything worthwhile to say at all to add nuance.
English
2
0
5
266
trey retweetledi
Andy Masley
Andy Masley@AndyMasley·
I've had a very nice back-and-forth email exchange with Shaolei Ren, who provided the original "AI uses a bottle of water per prompt" estimate with the Washington Post. I wanna be clear that he's more concerned than I am about AI's broader water use, so I don't want this to be read as him agreeing with me on my larger points, but he specifically agrees that his estimate for the water bottle per prompt is now outdated. Specifically: "The 2024 estimate was time-specific, assumption-based, and should not be used to describe general AI/ChatGPT or today’s optimized systems." His specific estimate for an average chatbot's water cost now is ~15 mL if you include the offsite costs. 33x lower than the bottle of water. So I do want to plant a flag and say "The person who provided the original sole estimate that chatbots use a bottle of water per prompt agrees that now that we have better information and more optimized systems, this is no longer the case, and models use ~15 mL each if you include the offsite costs, roughly 5 mL in the data center itself." It's silly that in 2026 educated people still believe that AI uses a bottle of water per normal prompt. No one investigating this believes it anymore.
English
8
26
249
8.2K
trey retweetledi
Edward White
Edward White@edwardwhitenz·
BYD for years used billions owed to suppliers as cheap financing to fund its rapid growth from a little-known component maker to a powerhouse reshaping the world’s auto industry. Now, Beijing is forcing it to clean up its books: ft.com/content/336bf3…
Edward White tweet media
English
10
30
99
18K
trey retweetledi
Enguerrand VII de Coucy
Enguerrand VII de Coucy@ingelramdecoucy·
“At least the Chinese won’t copy that car” has to be among the most brutal things this man can say about the Ferrari abomination. Somebody call the burn ward
Antonello Guerrera@antoguerrera

Former Ferrari chairman Montezemolo tears the new electric Ferrari “Luce” apart: “I cannot say what I really think: I would harm Ferrari. We risk the destruction of a legend. So sorry. Take the Prancing Horse off. At least the Chinese won’t copy this car”

English
70
1K
13.9K
407.1K
trey retweetledi
joseph francis
joseph francis@joefrancis505·
AJR showed that it is possible to win a Nobel Prize using Mickey Mouse Numbers. In my new blog post, I show how their results depend on choice made in data construction, in both "The Colonial Origins of Comparative Development" (2001) and "Reversal of Fortune" (2002). 1/9
joseph francis tweet media
English
8
32
155
16.5K
trey retweetledi
trey retweetledi
Senior PowerPoint Engineer
I believe this but it's also probably a problem with software in general? Uber has north of 2,000 software engineers, what do you suppose they all doing all day? Is Uber 500 engineering labor-years better today than it was 3 months ago?
Ed Zitron@edzitron

Uber’s COO has said that it’s getting “harder to justify” its AI costs because there was no way to show a link between AI spend and any meaningful increase in useful features. This is the first time I’ve seen a company say this directly. businessinsider.com/uber-coo-andre…

English
34
13
721
90.7K
trey retweetledi
Boris Cherny
Boris Cherny@bcherny·
> … [W]e keep finding things that are mysterious, even unsettling. We find structures that mirror results from human neuroscience. We find evidence of introspection. We find internal states that functionally mirror joy, satisfaction, fear, grief, and unease. I don’t know what that means, but I think it warrants ongoing discernment. > We need more of the world—religious communities, civil society, scholars, governments, and indeed all people of good will … to take this seriously, to look closely, and to push events in a better direction. We need informed critics who will tell the labs when we are failing. We need moral voices that the incentives cannot bend.
Anthropic@AnthropicAI

Anthropic co-founder Chris Olah was invited to speak at today's presentation of Pope Leo XIV's encyclical "Magnifica humanitas." Read the full text of his remarks: anthropic.com/news/chris-ola…

English
237
175
2.2K
317.6K
trey
trey@Comparativist·
@gbrl_dick @Miles_Brundage as to the moral panic: it’s treating every type of AI assisted writing as the equivalent of a “ChatGPT, write an essay about X” like slopmills do. Was it Stanford that just banned even AI assisted brainstorming for law students? That’s literally crazy.
English
0
0
4
60
trey
trey@Comparativist·
@gbrl_dick @Miles_Brundage for me… just looks like a moral panic around AI generated text. And I’ve seen plenty of false positive examples and know decent prompting gets around it.
English
3
0
6
317
trey retweetledi
Miles Brundage
Miles Brundage@Miles_Brundage·
Many of you are vastly overconfident in Pangram
English
29
21
432
35.1K
trey
trey@Comparativist·
very cool.
Kvali@kvali_app

Everyone talks about Denisovan DNA in Tibet, but the real mystery has always been the other "ghost" population hiding in the Tibetan genome. Archaeogeneticists finally tracked them down, and they did not originate on the high plateau. For years, ancient DNA models showed that Tibetans are a product of a profound genetic fusion. About 80% of their ancestry comes from Neolithic Yellow River millet farmers in Northeast China who expanded upward thousands of years ago. The remaining 20% was a total mystery: an uncharacterized ghost lineage that did not match any known ancient skeleton. A major study led by Fu Qiaomei solved this by sequencing a 7,100-year-old hunter-gatherer from the Xingyi site in Yunnan. This individual represents a deeply basal Asian branch that split from other East Asians over 40,000 years ago. These Paleolithic survivors weathered the Ice Age in the rugged valleys of southwest China. When they eventually migrated onto the plateau and mixed with northern farmers, they created the distinct ancestral baseline of modern Tibetans. Here is how the ancestral makeup breaks down: 🧬 80% Northern East Asian (Yellow River agricultural baseline) ⛰️ 20% Basal Asian "Xingyi" ghost lineage (Paleolithic survivors) 🫁 Early stabilization of the Denisovan-derived EPAS1 gene for hypoxia The Tibetan genome represents a highly distinct genetic crossroads, rewriting the old theory that the plateau was populated by a simple, late migration of lowland farmers.

English
0
0
1
153
trey retweetledi
Delicious Tacos
Delicious Tacos@Delicious_Tacos·
“It’s my honor to announce that the true pope is working with us from Avignon”
Delicious Tacos tweet media
English
60
964
11.6K
377.3K
trey
trey@Comparativist·
Idk what google is doing. I get a freebie Gemini Pro subscription because I use so many other Google services and I find it a nearly unusable, stupid, base model. Terrible combination of being sycophantic and not following directions.
Replica@John82924749

🤑At the moment, the strongest Gemini 3.5 Pro checkpoint seems to outperform even some GPT-5.6 xHigh variants in logical reasoning and inverse reasoning, so I’m not sure whether they’ll position it directly against GPT-5.6 Pro.

English
1
0
2
228