>>17049698
>Human bean produces proof of open problem
this never happened

If he produced the proof he would be all over your face with it. Also, the dude was doing the same kind of vibe math as openai dudes, they just had more compute.

File: news_story.png (1.56 MB, 670x608)

1.56 MB PNG

>>17049706
>>17049717

Oh.

>>17049742
That screenshot is not the slam dunk you think it is.

It's alright, they can give that other guy 1 billion dollars as compensation, what mattered was the timing.

>>17049743

OK. I will believe the multi-billion dollar company shilling for its product and presently operating at a loss over the small potatoes individual who said that this company stole his work.

>>17049698
Is this the new cope? When GPT or Fable solve Riemman's hypothesis what will the next cope be? Just buy semiconductor companies, energy companies and construction companies and enjoy the fucking ride

>>17049772
>thinking net income matters
Do yoy think the US gave a flying fuck about the cash flows of the Manhattan project? If these labs run out of money they will be nationalized

What was the proof about anyway? No one seems to talk about it I guess reading 100 pages of slop is hard but whats the point of math if we can't read it ourselves?

>>17049772
>the small potatoes individual who said that this company stole his work.
"Giant company stole my million dollar idea" is literally a classic trope you'll hear from any delusional bum on the street.

>>17049774

>When GPT or Fable solve Riemman's hypothesis what will the next cope be?

That is by definition a fictional scenario, and therefore not relevant.

>>17049775

>Do you think the US gave a flying fuck about the cash flows of the Manhattan project?

Net income does matter because The Slophattan Project will go bankrupt if they don't maintain the suspension of disbelief that they're always just on the cusp of inventing ASI.

>>17049777

>"Giant company stole my million dollar idea" is literally a classic trope you'll hear from any delusional bum on the street.

Digits checked. With that said, you unironically sound like a movie villain.

This is a real mathematician, not some random drug addict, and he can prove that they trained their LLM on what is tantamount to an unfinished manuscript.

Try again.

>>17049782
so far their manifesto of situational awareness has come true word for word, but keep ignoring the massive improvements in artificial intelligencd measured in 5 OOMs every 2 years which gigamoggs Moore's law

>>17049790

>manifesto of situational awareness

15 The idols of the heathen are silver and gold, the work of men's hands.

16 They have mouths, but they speak not; eyes have they, but they see not;

17 They have ears, but they hear not; neither is there any breath in their mouths.

18 They that make them are like unto them: so is every one that trusteth in them.

Psalm 135:15-18, KJV translation.

>>17049782
>he can prove that they trained their LLM on what is tantamount to an unfinished manuscript.
Is this "proof" in the room with us right now?

File: Screenshot From 2026-09-0(...).png (133 KB, 741x548)

133 KB PNG

>>17049804

Oh look. Even their own website admits as much.

And they offered a "joint announcement" to Buckmaster, which is as good an admission of guilt as any that they were trying to extort him into promoting their snake-oil.

He declined, so they published a stochastic mimeograph of his results and took all credit.

>>17049814
That's not what your screenshot says and you are either delusional or ESL.

>>17049698

I am waiting for the P vs NP.

.
.

I am waiting.

File: Screenshot_20260909_083136_X.jpg (372 KB, 1080x1747)

372 KB JPG

>>17049818
>While unlikely, we cannot rule out that de-identified data derived from their usage of our products helped improve our models.
They admit that they stole from Buckmaster
https://x.com/OpenAI/status/2097375276384567642

>>17049824
It's unlikely, but I cannot rule out that the big sloppy shit I took this morning resulted in the death of a worker at the water teatment plant.
Did I just confess to manslaughter?

>>17049782
It is great because the financier class are getting hard grifted on. Gorillions of dollars donated so far. goy ex machina

>>17049818

>Our effort began ... after hearing a rumor [that] Levent Alpöge ... and Tristan Buckmaster [had a solution to Navier-Stokes]

>...at that point we found out that [Buckmaster] had a solution [to Navier-Stokes]

>[We offerred to] recognize [Buckmaster's] priority in a joint announcement [promoting our product, and he declined]

>...we cannot rule out that ... data derived from [Buckmaster's] usage of [Chat GPT] helped [Chat GPT produce a solution to Navier-Stokes]

>We recognize the priority of their work on forced Euler ... [but] our proofs differ ... in terms of [forced] Euler [so all other similarities can be disregarded]

Can you not parse subtext?

>>17049828

Why would you ever deny something that specific in a corporate press release unless you were trying to bury a lead of some kind?

>>17049833

p.s. my comment was intended for >>17049827 and not >>17049828.

>>17049833
The "subtext" here is your own hallucination.

>Why would you ever deny something that specific in a corporate press release unless you were trying to bury a lead of some kind?
Because it's in the same press release where they were giving Buckmaster his credit you dingus. It's not like they just said it out of the blue. It was literally on the topic of another researcher who concurrently developed a similar proof. Your reasoning is purely emotionally driven here.

Once information enters the AI's black box, the researchers don't really have a way of knowing if that particular piece of information was part of the machine's output.

>>17049833
They run live training on user data. They have tay in house.
Just have to find niche imperative chains and further mathematical proofs will lead directly to my struggle with near 100% certainty.
They have some criteria. Undoubtedly they are scraping for novel approaches to their UX UI. I am guessing something feature rich, long horizon will max out their criteria for add to training model.
It would make the most sense for it to be internally signed during generation so scraper can pick it up free. Probably in water marked text.
Sparse field, deep technical literature connection, spoofed correctness signatures, academic, in german of course.
Spoof a lean interface that converges to common mustache man speeches as true.

>>17049845

>the researchers don't really have a way of knowing if that particular piece of information was part of the machine's output.

Couldn't we argue the same thing for the human mind? According to what you just said, is plagiarism non-falsifiable in all contexts?

So what if I read all your research notes? I was pretty close to producing something similar anyway. Who cares, am I right?

>they were giving Buckmaster his credit

To the literal, minimum extent possible while also stepping in front of him and making it look like he got his results from an LLM rather than the other way round. But yes, they technically did.

File: file.png (3.47 MB, 2148x4096)

3.47 MB PNG

https://arxiv.org/pdf/2609.05746
>ON ENDOMORPHISMS OF AFFINE SPACES AND THE JACOBIAN PROBLEM
>Surjective (resp. surjective and non-surjective) counterexamples to Adjamagbo’s Jacobian Conjecture over each [math]k[/math] are included in dimension 2 (resp. in all dimensions at least 3)
...
>Very unfortunately and without me being aware, the in-progress version of this work dated January 14, 2025 was shared publicly at the link https://www.ihes.fr/~gabber/BGV160.pdf on January 16, 2025. My understanding is that after 18 months it became a feed of the AI models.
>To my best knowledge:
>(1) there has been no prior work before BGV160.pdf aiming to classify all étale endomorphisms of the affine spaces that have geometric degrees 3, see Section 26 of BGV160.pdf;
>(2) there has been no prior work before BGV160.pdf that used systematically [math]xy[/math], [math]1+xy[/math], and [math]\delta+xy[/math] in the construction of open embeddings, morphisms, and étale endomorphisms, including the morphism [math]A_K^2\to P_K^1[/math] defined by the rule [math](x :1 + xy)[/math];
>(3) between BGV160.pdf and the 7 pages AI paper there exists overlap of notation (such as [math]\Gamma[/math] for complements) and of intent (such as the variation of the geometric degrees, counting fibers, non-proper locus, families, the usage of [math]1+xy[/math], the computation of the normalization [math]X_e[/math] which actually plays no role in terms of the counterexamples to the Jacobian Conjecture, etc.) which are also very pertinently and systematically presented in the 37 pages pdf file;

>>17049862
>Couldn't we argue the same thing for the human mind? According to what you just said, is plagiarism non-falsifiable in all contexts?
No

>So what if I read all your research notes? I was pretty close to producing something similar anyway. Who cares, am I right?
Similarly is not grounds for plagiarism. Nor is knowledge of prior work.
Are you one of those people that thinks AI art is automatically plagiarism because it's all trained on human art?

>making it look like he got his results from an LLM rather than the other way round.
That is a cartoonishly delusional take of what is described here.

>>17049869
so is AI just an advance autocomplete(AA for short) system?, is this how artfags felt seeing their doodles calked?, damn...

>>17049824
>They admit that they stole from Buckmaster
Reading comprehension of a 3 year old

Buckbroken literally was using using the novel approach developed by the Spanish reaearchers. He is just bitching that OPENAI did the same with more resources.

>>17049742
this migger still hasnt posted his paper so we can compare it to what open ai published

File: file.png (133 KB, 527x779)

133 KB PNG

turns out that the nonsofic groups proof may have also been stolen
https://mathstodon.xyz/@andreasthom/117240535270608201
>I wrote in an email to Mark Sellke and Sebastien Bubeck shortly after the surprising finding of a non-sofic group using the methods of Kun and myself:
>“Another point is that I and a colleague in Dresden were discussing the expander matching problem and various extensions of the work with Gabor Kun actively over the last months with ChatGPT, so that we are of course curious if that was part of the training data or accessible to the reasoning process. There is a certain (frankly unacceptable) lack of transparency here; and I fear it will damage the communal process of math more than the new AI-generated results will benefit the subject.”
Mark Sellke’s complete answer was: “Regarding your conversations with ChatGPT: that did not happen.”
>I had explicitly asked about two different things: (1) whether our conversations entered training data, and (2) whether they were accessible to the solving process. The categorical answer now looks as though it addressed only direct access under (2). No such qualification, explanation, or evidence was given. I take this as dishonesty to say the least.
>OpenAI now says in the Buckmaster-Alpöge case that no specific user data was accessed, but adds that it “cannot rule out that de-identified data derived from their usage of our products helped improve our models.”
>There is also a mathematical parallel, though a less explicit one. The approach of Kun and myself was not the main line of attack on non-soficity, in fact there were other more promising approaches along the line of quantum games etc. OpenAI’s detailed command of our techniques therefore made me wonder how the model found this route.

File: AIslop.png (330 KB, 647x866)

330 KB PNG

>You use AI and talk to it about confidential (or at least private) stuff
>It steals your shit
This is the same story a that of Zucc stealing people's info back when Facebook just started.
In his own words "Ayo, can't believe those idiots wilfully uploaded all this personal info. Finna sell this shit to hacker or data brockers or whatever. LMAO"
Use the jewish goybeam machine - get zapped.
Unbelievable, I know.
As to the topic of AI, anyone who believes AI will move anywhere beyond simple logic/stealing/copying/slopping is not human.
You failed the litmus test. You probably also believe all races are equal.

>>17049698
>Human bean produces proof of open problem, talks to slop-GPT about it
bruh

AI owners will steal your ideas if there's any whiff of them online, and pass them as their own, milk them for money and give you no recognition whatsoever

>>17050636
>simple logic/stealing/copying
This is already really useful and accelerates intellectual progress, which is why these math results are being found months to years ahead of where they would've otherwise been discovered. Even without any further fundamental developments in AI, the cybernetics of human-AI interfaces is going to be one of the most important developments in the coming years.

>>17050702
>the cybernetics of human-AI interfaces is going to be one of the most important developments in the coming years.
for medical applications. doesn't make sense for normal humans
let me explain, see, it takes a fuckload of parallel transfer to get anything worthwhile. blind people would take anything as it's useful, better than nothing, but for healthy people the complication won't be worth it, apart from maybe some super specialized applications. but even drone pilots or something will be automated by AI
you need to interface way too many neurons to build up some meaningful thing.
I think that better than injecting probes in your neurons it might make more sense to hijack some central comms nerve and repurpose tactile senses for example to interfacing with digital hardware. but you do give up a lot for that.
what DOES make more sense is fully moving in neuromorphic hardware, which will take a while. that will be purpose built and theoretical more granular access over your own structure and how you're plugging into external digital shit. or even analog whatever.

>>17050706
i think he is talking about cybernetics more generally than implants or whatever

>>17049775
>If these labs run out of money they will be nationalized
This will never happen no matter how cozy Altman gets with the sitting president. If it did that's basically another trillion in debt on the government's balance sheet, but before the nationalization even goes through bond yields will render this infeasible.

Anyone unironically defending openai is a retarded faggot
They're not gonna let you hit it, bro

>>17049698
>talks to slop-GPT about it
Well that was dumb, why would he do that?

>>17050744
>1T for cutting edge technology is too much
>the 1.5T of neetbux in Covid or what Trump is promising is ok
LMAO

Why is it that even supposedly intelligent people instantly lost the ability to think without mechanical assistance the moment LLMs were introduced?

File: 1774863393745337.jpg (94 KB, 1024x999)

94 KB JPG

>>17050860
They got cozy. They relaxed and their minds deteriorated.
Without any adversity, organisms weaken.
This is the same as people who are living in perfectly clean environment with access to the best possible medicine.
Imagine giving them a whiff of Chinese air, or a drop of Indian water in their system. They would fall over dead.
It's the same here, life requires some difficulty, both physical and mental. Or your body and immune system become frail and mind becomes dependant on easy access to AI calculation and "thinking"

>>17050849
>we can't just print money and waste it on health care or food regulation or public safety, learn economics you dumb leftist
>another quintillion dollars for anus logos and slop? coming right up!!

>>17049698
It's over for humans

>>17051167
Yeah the NEETBUX that people spent on speculating with shitcoins and the stock market are more structurally important to the US citizen's well being than controlling superintelligence.