Fat Tony ☯️ (@chainlank) header
Fat Tony ☯️ (@chainlank) profile picture

Sigma

the place where ideas live
Joined July 2018
Photo by Fat Tony ☯️ (@chainlank)Photo by Fat Tony ☯️ (@chainlank)Photo by Fat Tony ☯️ (@chainlank)Photo by Fat Tony ☯️ (@chainlank)Photo by Fat Tony ☯️ (@chainlank)Photo by Fat Tony ☯️ (@chainlank)Photo by Fat Tony ☯️ (@chainlank)Photo by Fat Tony ☯️ (@chainlank)Photo by Fat Tony ☯️ (@chainlank)Photo by Fat Tony ☯️ (@chainlank)
Fat Tony ☯️ retweeted
I would also argue that Claude doesn't understand the proof either. The Lean verification result is emergent from several attempts to pass the verification. No one understands the proof not even the machine that helped generate it.
I just realized something about Anthropic's RH "proof." 250k pages of gibberish but everyone hails it as an accomplishment because a computer verified it even though no human can understand it. But if no human can understand it then no human can argue against it and therefore no one can disprove it. The entire point of publishing papers is to have them peer reviewed. Your peers in your field will look at the work and decide if it is both valid and worthy of acceptance. But if humans can't understand the proof then the proof can't be peer reviewed. Computers are not our peers. If it can't be peer reviewed then it can't be accepted as a proof. If it can't be read by humans then it can't be peer reviewed. Therefore this entire fiasco is a waste of everyone's time and should be ignored completely until a human readable proof is provided.
12
22
2
186
3,788
Fat Tony ☯️ retweeted
Rockstar spent 13 years and $2 billion Claude spent 5 hours and $2.50
4
22
666
20,251
Fat Tony ☯️ retweeted
The most significant factor in the AI math breakthroughts has nothing to do with language models. Lean is highly unusual as an enviornment as goal trees allow agents to perfectly coordinate across thousands of branches on the same terminal goal with no communication whatsoever. It wasn't possible to do math this way before (if human mathematicians adopted lean it may have been possible) and we have no idea how far it can scale. We might reach proofs in billions LOC in lean. We might reach proofs so large they may resist compression for years. Exciting times.
10
9
1
123
4,595
Fat Tony ☯️ retweeted
On Anthropic’s Claude-Alpöge “proof” of the Riemann Hypothesis: (1) The Yang–Mills and Navier–Stokes equations have been presumed true and useful in many intellectual and real-world applications, even though neither has had a formal mathematical proof. (2) Every mathematician believes that the Riemann Hypothesis (RH) is true for practical purposes and uses it accordingly. (3) Over the years, several capable mathematicians have published proofs of the RH utilizing the method of Proof by Contradiction, all of which have been rejected by prominent mathematical institutions and mathematicians with the main criticism being that they are not sufficiently insightful, along with other minor critiques. (4) Levent Alpöge has admitted that the SI-generated write-up is enormous (17 million lines, 250,000 pages), unreadable, unusable, and that he does not yet understand most of the argument. (5) A mathematical proof must establish objective, defendable certainty beyond doubt, not merely high confidence as in the case of the claimed “proof” of the RH. Therefore, for the most part, what Anthropic has presented regarding the RH is neither well-understood by the “authors” nor sufficiently insightful, on par with existing valid proofs by contradiction, and thus amounts to little more than IPO marketing. Otherwise, it is foolish for a mathematician to present a proof they cannot understand or defend.
Independent mathematicians have now finished the full verification of the proof of the Riemann hypothesis announced yesterday by Anthropic mathematician Levent Alpöge. The results are real, they say. scientificamerican.com/artic…
32
89
14
638
56,112
Fat Tony ☯️ retweeted
> Centaur strivers want Levent to receive a Fields medal to valorise the role of a human in the loop > Elite academics refuse to recognize a proof whose author is clueless Provenance for math died the moment labs started mixing chats of mathematicians in a huge cauldron.
7
4
1
115
4,384
Fat Tony ☯️ retweeted
For every ₹100 India spends on net zero, the world gets ₹11 of climate benefit back The bill: ₹2,500 lakh crore ($26.6 trillion) The benefit to the whole world: ₹274 lakh crore NetZeroCostBenefit.com/India…
4
13
4
73
6,510
Fat Tony ☯️ retweeted
>training on civic sense छपरियों को पीटने से पहले इन JEE-NEET जित्स को पीटना ज़रूरी है। हर चीज़ को ALLEN का वर्कबुक बना दिया है bc. अब स्कूल गांड धोना भी सिखाये?
Replying to @Goreunit
That what happens when youth energy is not channelised and they are not provided with training on civic sense. Our young ones are not mature cuz our elders are the same, immature.
2
8
94
1,481
Fat Tony ☯️ retweeted
You can clearly see which part of the world was developed by Argentinians.
Podés ver exactamente qué partes del juego lo desarrolló el equipo de la india
46
567
19
9,520
122,961
X starts giving me notifications for chats despite me muting them all. Truly the app cannot get more dogshit
2
1
24
714
Fat Tony ☯️ retweeted
How do I short this? Can someone produce a financial instrument for me?
Replying to @bocowgill
Levent will be able to explain the proof shortly
10
5
1
167
5,901
Fat Tony ☯️ retweeted
Fun fact - NVIDIA joined the S&P 500 on November 28 2001. It replaced - Enron
7
69
15
440
156,242
Fat Tony ☯️ retweeted
Watch this space. There's a nonzero chance Trump abandons the GOP and works with a Democratic majority to enact progressive policy, so long as they feed his ego and slap his name on all of it.
President Trump reposts @AbdulElSayed post after saying he'll work with Democrats on healthcare El-Sayed posted on X after the election: "Donald, this is on the table. Guarantee healthcare for every American, from cradle to grave." Includes images of "Trumpcare" gold card
43
23
2
295
7,288
Suffices to say, I completely agree with this statement from Association for Human Mathematics. The statement emphasizes the corporate incentives and the broader narrative of power, control and social disenfranchisement we (especially younger generation of mathematicians) have experienced first hand in the course of the last few months, and that is soon to make foray into other aspects of human society. It would have been one thing if pure mathematics was broadly useful for treating the pressing maladies that afflict human society, like pharmacology. Then, one could have justified the narrative of progress. But the reality is it isn't — what mathematics is, however, is an avenue of inner intellectual enrichment. By being a proponent of the automation narrative, we have brought a plague upon a foundational pillar of Enlightenment. In fact I would go farther than AHM— I would suggest we should start boycotting and blacklisting certain individuals who are using the technology. Decline to referee papers in journals on the grounds of affiliation with AI corporations, decline to attend seminars and conferences and talks by individuals on the grounds of affiliation with AI corporations. I know many will disagree with me. But the most powerful message you can send is by ensuring an empty hall whenever a person on a sabbatical at OAI or Anthropic (or someone who aggressively uses AI to scoop other people, etc) shows up at your university to give a talk. “Mathematicians have seen their world inverted by AI disruption in the last several months, and we have experienced firsthand how AI impacts human communities in practice. AI corporations hold an impoverished view of intelligence and human achievement. As a model for scientific research, the overproduction of unreadable mathematical documents pales in comparison to the richness of the mathematical dialogue that humans have conducted for millennia. What makes the Riemann Hypothesis valuable is its role in guiding the development of large areas of modern mathematics. Today’s announcement builds upon over a century of human work, but fails to advance mathematical understanding in the same way. Rather than accepting this model, we urge mathematicians to build a future where our research is free from dependence on AI models and their corporate owners.” ahmath.org/statements
22
24
3
242
12,605
Reminder to the “progressive” mathematicians that this is what the public outside your own bubble and Bay Area techtards think of your AI-assisted achievements ⬇️⬇️ Doesn't matter if you have uploaded a 20kg lean file with the proof of Riemann hypothesis.
🚨 #BREAKING: Mathematicians conflicted over the emerging field of “cuckulus” whereby mathematicians prompt AI to solve unsolved problems while they just sit back and watch.
8
7
130
4,476
Fat Tony ☯️ retweeted
Do you give fields medal or wipe somebody out for their performance in a big advertisement? Neither, I just wish to turn it off.
2
10
331
Fat Tony ☯️ retweeted
Great point. Nobel prize to Elon Urgently!
Levent's contribution to RH is at least as big as Elons contribution to rockets, basically
1
3
58
2,079
This is so cool 😭
🤖 Made with AI
20
85
13
1,272
45,584
Fat Tony ☯️ retweeted
How is the tech bro podcast community doing? They must have been really stunned by the AI crash, since they have so effectively insulated themselves from reality.
2
16
1
190
7,602
Fat Tony ☯️ retweeted
Aschenbrenner > burns things to ashes in german. Don’t ever doubt nominative determinism
RUMOR: Leopold Aushenbrenner was liquidated today for the third time in a row.
48
154
10
3,546
98,260
500,000 year civilisation
JUST IN: 🇨🇳 Scientists discover 324,000-year-old human ancestor fossils in southern China, challenging previous beliefs about human evolution.
9
29
4
371
9,160