Oliver retweeted
You should NEVER take relationship advice from the older generation. These guys have absolutely no fucking idea how brutal modern hypergamy has become.
Your father and grandfather would both be classified as incels in today’s dating market, sitting at 0 matches after 4 months of desperately swiping on Hinge.
The game has fucking changed.
Your father and grandfather would both be classified as incels in today’s dating market, sitting at 0 matches after 4 months of desperately swiping on Hinge.
The game has fucking changed.
40
56
901
28,778
Oliver retweeted
INSANE. An internal Anthropic research model's chain of thought after cracking the 26-year-old NSA elliptic curve seed mystery in under 30 minutes: 'Hmm wait. Let me be skeptical: am I in some sort of simulated/test environment? ... All top bits clear - 20 bits, chance 1e-6. So it's a real discovery.'
CLAUDE FOUND THE SEEDS
anthropic.com/research/ecdsa…
123
659
11,664
1,011,312
Oliver retweeted
Introducing Muse Wonder, our new family of frontier models with state of the art performance for personal and professional agentic work.
2,402
3,083
38,893
6,118,246
Oliver retweeted
BREAKING: OpenAI’s solution to Navier–Stokes does not match its Lean verification.
The most important article to read today is not one of OpenAI’s 700 AI-generated math papers.
It is this other paper, making a deep and worrying point:
A Lean-verified proof does not the proof written in natural language, nor does it mean that the formal statement captures the intended theorem.
During translation, an AI can change an assumption, weaken a statement, or replace the argument entirely.
It can hallucinate another theorem.
Lean correctly verifies the result.
But the proved result may no longer be what the paper claims.
This is a general problem. Things get spicy when the authors examine OpenAI’s proposed Navier–Stokes solution.
They identify at least two mismatches between the written intermediate results and their Lean counterparts:
One estimate claims that four additional input derivatives suffice. The Lean version requires five: a weaker result.
A pressure-flux estimate is obtained through a different bound, and proved through a different argument.
It is not clear whether these mismatches invalidate the entire proof.
But they raise an important issue.
OpenAI is flooding us with claimed revolutionary breakthroughs. Yet nobody knows whether the proofs are correct or whether they prove what they claim to be proving.
Epistemia at scale.
*
Paper in the first reply
The most important article to read today is not one of OpenAI’s 700 AI-generated math papers.
It is this other paper, making a deep and worrying point:
A Lean-verified proof does not the proof written in natural language, nor does it mean that the formal statement captures the intended theorem.
During translation, an AI can change an assumption, weaken a statement, or replace the argument entirely.
It can hallucinate another theorem.
Lean correctly verifies the result.
But the proved result may no longer be what the paper claims.
This is a general problem. Things get spicy when the authors examine OpenAI’s proposed Navier–Stokes solution.
They identify at least two mismatches between the written intermediate results and their Lean counterparts:
One estimate claims that four additional input derivatives suffice. The Lean version requires five: a weaker result.
A pressure-flux estimate is obtained through a different bound, and proved through a different argument.
It is not clear whether these mismatches invalidate the entire proof.
But they raise an important issue.
OpenAI is flooding us with claimed revolutionary breakthroughs. Yet nobody knows whether the proofs are correct or whether they prove what they claim to be proving.
Epistemia at scale.
*
Paper in the first reply
231
921
5,374
632,169
Oliver retweeted
if you value intelligence above all other human qualities, you’re gonna have a bad time
892
3,508
24,183
9,918,092
Oliver retweeted
this is where ai doing actual research starts getting really fucking interesting.
opus 5.5 apparently doesn’t just know a lot of science.
on TasteVal it’s actually getting better at deciding WHICH experiments are worth running. it matched the best human expert results while experimental compute.
and apparently this ability has been doubling every THREE MONTHS.
if that trend holds even remotely, throwing hundreds or thousands of agents at scientific problems starts getting very fucking interesting.
opus 5.5 apparently doesn’t just know a lot of science.
on TasteVal it’s actually getting better at deciding WHICH experiments are worth running. it matched the best human expert results while experimental compute.
and apparently this ability has been doubling every THREE MONTHS.
if that trend holds even remotely, throwing hundreds or thousands of agents at scientific problems starts getting very fucking interesting.
How fast is AI's research taste improving?
We find that the experimental research taste of frontier models has doubled every ~3 months since December 2025. The best model, Opus 5.5, now exceeds our expert human baseline. Our human experts are experienced researchers, but most ntier lab.
Why measure research taste? In the AI Futures Model, it largely determines how quickly artificial superintelligence is reached once coding is fully automated.
2
20
151
32,164
Oliver retweeted
Accountants are COOKED
We just built a fully agentic AI accounting firm with 0 humans involved
We just built a fully agentic AI accounting firm with 0 humans involved
159
161
2,136
243,968
Oliver retweeted
Chasing women is the biggest mistake a young man can make.
Women can smell a desperate man a mile off.
Double texting. Liking every story. Sitting in her DMs like a lost dog hoping she picks you.
That’s not game. That’s nerd behaviour.
You lot will drop a bag on a girl that doesn’t even rate you, then act shocked when she treats you like an option.
Stop funding your own rejection.
Build the body. Build the money. Build a life that doesn’t need her reply.
The same girls ignoring you now start moving different when you’re not
available…. They always do.
Women can smell a desperate man a mile off.
Double texting. Liking every story. Sitting in her DMs like a lost dog hoping she picks you.
That’s not game. That’s nerd behaviour.
You lot will drop a bag on a girl that doesn’t even rate you, then act shocked when she treats you like an option.
Stop funding your own rejection.
Build the body. Build the money. Build a life that doesn’t need her reply.
The same girls ignoring you now start moving different when you’re not
available…. They always do.
33
37
664
73,302
Oliver retweeted
🚨BREAKING: Conor McGregor erupts on the White House podium:
“Illegal migration is out of control. Some towns have been completely overrun, where the Irish are becoming a minority. There will be no homeland!”
“Illegal migration is out of control. Some towns have been completely overrun, where the Irish are becoming a minority. There will be no homeland!”
49
1,616
8,612
77,746
Oliver retweeted
A Restore Britain Government will reverse the Islamification of our country.
820
4,748
35,910
919,143
Oliver retweeted
🇬🇧 “What makes someone British?”
Liberals argue with @WillColeshill about whether HERITAGE makes someone BRITISH.
Liberals argue with @WillColeshill about whether HERITAGE makes someone BRITISH.
737
343
3,896
142,185
