OpenAI said its new AI agents solved a 90-year-old math problem, known as the Navier-Stokes equations, in under four days. Now, an NYU professor is speaking out, claiming the AI ‘piggy-backed’ off of his research.

Like humans. We all piggyback off of those who preceded us.

@Teh_Snowflake I wrote the greatest song ever by stealing other people’s music.

That’s exactly how discovery is supposed to work. What a maroon.

Yesterday's news now I guess, onto mass hysteria over a problem anyone with brain knew about 30 years ago.

So OpenAI claims a breakthrough on Navier-Stokes in four days, but an NYU prof says they just leaned on his work? That's some messy academic credit situation right there.

Is piggy-backed off the opposite of piggy-backed on? Are on and off now no longer antonyms?

What does this mean for medical research, technology development, or anything someone is working on when a lab can step in, take your idea, and assign an army of agents and compute to claim a patent first?

If he couldn't piggy back off his own work then what's he bitching about?

Why is it taking anyone 90 years to solve a math problem?

Basically, someone worked their ass off trying to solve the problem but an AI agent got the information, analyzed the problem and solved it

Bullshit this is stolen work!

The people in the comments just love thievery by white Zionists

According to the OpenAi document, the external force definition as the residual is by construction and therefore in time in a discrete sense. However, the limiting sense as t approaches 1 is not finite and therefore the resulting pde analyzed is not the one by construction.

Cancer next ?

@Zigmanfreud Really doesn’t matter, the professor couldn’t figure it out. Soo

the real question is how much of the breakthrough was genuinely novel. That distinction matters a lot. 👀

Yeah maybe. 18 days ago. OpenAi lives my work. A lovingly parasitic relationship now for 2 years. All them where they got their architecture.

who's checked that?

Which will mean nothing once they use it to off us.

if using previous work is “piggy-backing” then every mathematician who ever cited a paper is a fraud. this is just cope at this point orz

The money and safety are in the solution

broader context on real world utility here:

That’s the real question: not just whether AI solved the problem, but whether the path to the solution was truly independent. If researchers’ work may have influenced the result, OpenAI should be transparent about exactly what data its agents accessed and how.

NYU dice que piggy-backearon. Da igual el crédito. El run de 88 horas ya tiene precio de 15 millones. El prize de uno nadie lo toca.