Once the models start to plateau - the frontier labs will probably start investing heavily into ASICing their most advanced
models.
Cheap inference which can be sold at a healthy margin, and also dogfooded internally to push the boundaries of the frontier itself.
1
1
112
ASIC is not discussed about enough
Everybody’s talking about training related advances.
But we literally have tech that will allow us to print a model on a chip and run it faster and cheaper.
What happens when you print a million Mythos’ on ASIC and run them in a DC?
1
115
“when they almost certainly got it right anyway”
if only there was a method, some mechanism, to catch the cases when they almost did not get it right. if only. we will never know.
i deleted all the tests in my repo and banned them in AGENTS.md
tests were always just a second, worse copy of your logic. a leaky imitation humans wrote because we couldn't hold every invariant in our head (staying green half the time the real bug happened anyway)
opus 5.5+ is good enough at raw logic now to read the intended invariants straight from the codebase
maintaining that duplicate copy is just economic waste. plus it makes tasks take way longer while agents burn compute running tests when they almost certainly got it right anyway
21
insane
In my ongoing quest to find the limits of Opus 5.5, I directed it to audit my old scientific field, theoretical astrophysics. We indexed about 150,000 papers, audited 50,000 for issues, and did progressively more thorough checks on thousands more. We found nearly 1000 major errors that got cited many times, most are still uncorrected. This survey enabled me to estimate the efficacy of peer review for correcting errors (not very good) and to identify dozens of open questions that anyone with spare Claude Max tokens could answer today (.md context files included).
Bottom line: AI can speed up quantitative science by >1000x with no further model improvement required. But I'm seeing no evidence of this being pushed within universities, only academics complaining about labs dropping math proofs.
Among my fellow reformed academic high agency founders, however, several are speed running entire disciplines.
This survey has implications for how science will be run in the very near future. Hint: Academic publishing is dead. It was already a very stupid way to record and promulgate knowledge, and had been so for decades, even before all that publicly funded knowledge was locked behind paywall cartels. Now AIs can generate thousands of novel, correct papers per day, it is no longer meaningful for humans to consider a research paper a "unit of science" or to even bother much with the 99.9% of science which, up to this point, was necessary drudgery. Time to discover something like General Relativity or Evolution every week. LFG!!
Enjoy: science.caseyhandmer.com/phy…
1
28
rex retweeted
Major IT systems start failing 18 months after drugged up silicon valley dipshits convince everyone to ditch their experienced coders for a machine that doesn't know how many Rs are in "strawberry".
Everyone: "Must be a cyber attack"
Bunch of pompous grifters.
.@paulfchristiano thinks that on the current trajectory (absent any slowing down), there is a 1 in 3 chance that we could have superintelligence powerful enough to overthrow the government within a year. (He articulates a concrete scenario).
He argues that the labs should pause further algorithmic progress until they can secure their weights and research secrets. Otherwise China could simply steal the superintelligence in the middle of the intelligence explosion, and the labs would no longer be able to slow down even if they saw their models going off the rails.
He also explains how AIs will be trained during the intelligence explosion, and why even a system trained entirely inside a data center could still have the capabilities profile such that humans would have very little chance of competing with it or controlling it.
And he discusses the plan for alignment - basically, making a superintelligence that actually does what we ask the way we mean it. Fixing reward hacks and building better graders will help up till the next few model generations, but he thinks that no single current approach has better than even odds of letting us control superintelligence. The hope is that the AIs figure out the solution to this problem during the intelligence explosion.
Paul is head of the Alignment Research Center. Previously, he was head of AI safety at the US AI Safety Institute. Before that, he led alignment at OpenAI, where he invented RLHF. He’s recently joined OpenAI’s board (though here he is speaking as a private citizen).
1
88
rex retweeted
The cultish valorisation of people who acquire certain narrow and lucrative kinds of technical expertise has blinded many to the fact that they are some of the dumbest, most clueless and insulated people in the world.
Anthropic Researcher Tao Lin: Pretty much all humans on earth will become permanently unemployable because of AI
"It's very likely that pretty much all humans on earth will become permanently unemployable"
"And will need to strictly rely on government welfare to sustain themselves"
"Human employment will go away."
Did my part. Please do yours frens.
Col Rohit Uniyal did us proud while battling the enemy. Now he is battling pancreatic cancer. Treatment cost (Daraxonrasib) is Rs 5 crore. His family has exhausted all their savings.
Let us fight this battle together. And win. I have done my little. Link: impactguru.com/s/JlOHio
1
97
rex retweeted
India's exports to ASEAN have grown by 58% in H1 2026-27. Interesting that exports to even Vietnam are growing quickly. Singapore and Malaysia are the largest ASEAN export markets for India.
pib.gov.in/PressReleaseDetai…
Whey protein causes kidney issues
Hamare dada pardada kabhi gym nahi gaye
Walking while exchaning chutkule and gossip with your boomer friends is an exercise
Housework is a replacement for strength training
What else did I miss
rex retweeted
Junction is bringing India’s best students together to solve ambitious semiconductor engineering problems.
Here’s a map of where every Junction applicant comes from🇮🇳
Two IIT Bombay students who built chip-making tools from an old projector and a drone motor have teamed up with an ace talent scout in Bengaluru to fly in about 15 students from across India and build semiconductor equipment in a fortnight.
@itsarnavb @aryamman_bhatia @abhinahiyaaar @hackerfabindia @_lagrangepoint
India's chip mission has money and fabs under construction. It is short of people who have run a fab tool and watched it misbehave.
The gap:
- ISM 2.0 outlay: Rs 1,27,500 crore across six pillars
- Tata's Dholera fab: 300 graduates training at Powerchip in Taiwan
- A 2 nm fab needs 200–500 engineers who have run one
- Japan's Rapidus: founded 2022, working 2 nm transistors by July 2025
-on seven decades of chipmaking
- Every tool in IIT Bombay's nanofab is imported
What the students built:
- HackerFab IITB, founded August 2025
- Lithography tool: ~Rs 3 lakh in materials (research-grade: ~Rs 1 crore)
- Full student fab: Rs 15–20 lakh
- First working devices on 12 June - a diode and a capacitor
- Prints 3–4 micrometre patterns, where IIT Bombay's first MOS transistors sat in the late 1980s
Junction, this December:
- ~15 students, two weeks, Lagrange Point in Indiranagar
- Flights, stay and project costs covered
- ~100 applications by launch night
- A 1980s tool that cost hundreds of thousands of dollars, rebuilt for $1,000–2,000
- Prototype turnaround: 1–2 days in China, 1–2 weeks in America, 1–2 months in India
What is still unproven:
- NMOS transistor expected in June, still listed as in progress
- No industry mentors named yet
- Hydrofluoric acid and 1,100°C furnaces, handled at speed in a house
Fair fucks to this man
Palestine APOLOGIZES to India 🇵🇸 🙏
The U-17 Head Coach Yousef Sharbaji condemned his Palestinian player's action, resulting in the Red card, while also congratulating India for the deserved victory 🇮🇳 👏 #IndianFootball
rex retweeted
Col Rohit Uniyal did us proud while battling the enemy. Now he is battling pancreatic cancer. Treatment cost (Daraxonrasib) is Rs 5 crore. His family has exhausted all their savings.
Let us fight this battle together. And win. I have done my little. Link: impactguru.com/s/JlOHio


































