Mom - don't let your sons grow up to be abstract (or computational) mathematicians.

https://www.scientificamerican.com/article/openai-unleashes-hundreds-more-math-results-upon-a-field-already-in-shock/

Hundreds more problems solved or progressed. Many already independently verified.

• Last time (Navier-Stokes progress) this took thousands of agents and million of dollars of compute
• This time - a single agent

Obviously, in aggregate, this is probably worth many Fields Medals. One in novel solution particular (the solution to the quasi-RH problem) has been commented on as being worthy of a Fields Medal on its own (if done by a human).

It appears SI is already dominating an area of reasoning and intelligence that many consider to be among the most difficult.

I foresee a huge drop in Math graduate program applications; how do you even begin to compete? Once again, glad to be an old. SI may open up a new world of prosperity, but it is going to be very bumpy.

More bold claims by OpenAI. Nothing of significance has been verified or replicated yet.

ntxVol said:

---

More bold claims by OpenAI. Nothing of significance has been verified or replicated yet.

---

Well, there is this if you bother to read the article.

"But many of the results have already been verified in Leana programming language that validates a proof's logicand are therefore all but certain to be correct."

Over_ed said:

---

ntxVol said:

---

More bold claims by OpenAI. Nothing of significance has been verified or replicated yet.

---

Well, there is this if you bother to read the article.

"But many of the results have already been verified in Leana programming language that validates a proof's logicand are therefore all but certain to be correct."

---

I did read the article, you cherry picked one snippet while the rest of the article full of doubt because OpenAI has made bold claims before.

Until mathematicians are able to test the model themselves and replicate the results, nobody is buying what OpenAI is selling.

They even put everything in github for all the skeptics to parse through.
.
https://openai.com/index/sharing-ai-progress-in-mathematics/

Now this is more brute force. The models still don't decide which problems are meaningful, they don't create the problems and abstractions that lead to the proofs. So a real mathematician that can do that is still ahead, but it's probably about 2 to 3 years until they can. BTW ther were 4000+ problems they tried to solve. Could only do around 700.

YouBet said:

---

I refuse to call it SI. It's AI until it solves scarcity.

---

Agree. I don't understand why Trump wants to change the name.

What does Cal Newport say about it though?!

ntxVol said:

---

More bold claims by OpenAI. Nothing of significance has been verified or replicated yet.

---

Many of the proposed proofs have been computer-validated using Lean, according to OpenAI. It reports about 42% of the manuscript-level main results are formalized, which is 300.

YouBet said:

---

I refuse to call it SI. It's AI until it solves scarcity.

---

Why? Actual intelligence hasn't solved scarcity.

Insisting on calling it Artificial Intelligence is like insisting that airplanes perform artificial flight.

GeorgiAg said:

---

YouBet said:

---

I refuse to call it SI. It's AI until it solves scarcity.

---

Agree. I don't understand why Trump wants to change the name.

---

Because AI was coined in a time when it was fictional. As we're actually living through the singularity now it's becoming obvious that it will quickly come up short in describing what this intelligence is.

TexAgs91 said:

---

YouBet said:

---

I refuse to call it SI. It's AI until it solves scarcity.

---

Why? Actual intelligence hasn't solved scarcity.

Insisting on calling it Artificial Intelligence is like insisting that airplanes perform artificial flight.

---

Exactly.

It's been known as AI since inception of the concept decades ago. It hasn't done anything to warrant a rebranding that suggests it's anything beyond even artificial.

I'm just saying to keep it as is until it solves something practical to warrant being called a Super Intelligence which implies something above and beyond sentience.

YouBet said:

---

TexAgs91 said:

---

YouBet said:

---

I refuse to call it SI. It's AI until it solves scarcity.

---

Why? Actual intelligence hasn't solved scarcity.

Insisting on calling it Artificial Intelligence is like insisting that airplanes perform artificial flight.

---

Exactly.

It's been known as AI since inception of the concept decades ago. It hasn't done anything to warrant a rebranding that suggests it's anything beyond even artificial.

I'm just saying to keep it as is until it solves something practical to warrant being called a Super Intelligence which implies something above and beyond sentience.

---

I've used it for things that would take a group of programmers a couple of weeks to do in 5 minutes. Holding out for a scarcity solution is moving the goalposts. This is intelligence... Wet or not.

JB99 said:

---

ntxVol said:

---

More bold claims by OpenAI. Nothing of significance has been verified or replicated yet.

---

They solved over 700 proofs that were verified. Try to read next time.

---

Anyone who places any value on OpenAI's claims is a fool. Nobody is going to believe them until they actually publish the model and allow mathematicians to replicate the results. If you actually read the article, you would know this.

Quote:

---

But OpenAI's reputation, among mathematicians, for bold claims and little transparency is causing considerable skepticism.

---

GeorgiAg said:

---

YouBet said:

---

I refuse to call it SI. It's AI until it solves scarcity.

---

Agree. I don't understand why Trump wants to change the name.

---

Trump changing the name is a money grab for Trump cronies….how?….all domain name variations of "super intelligence" were purchased before Trump's order

TexAgs91 said:

---

YouBet said:

---

TexAgs91 said:

---

YouBet said:

---

I refuse to call it SI. It's AI until it solves scarcity.

---

Why? Actual intelligence hasn't solved scarcity.

Insisting on calling it Artificial Intelligence is like insisting that airplanes perform artificial flight.

---

Exactly.

It's been known as AI since inception of the concept decades ago. It hasn't done anything to warrant a rebranding that suggests it's anything beyond even artificial.

I'm just saying to keep it as is until it solves something practical to warrant being called a Super Intelligence which implies something above and beyond sentience.

---

I've used it for things that would take a group of programmers a couple of weeks to do in 5 minutes. Holding out for a scarcity solution is moving the goalposts. This is intelligence... Wet or not.

---

I have as well and agree. Just saying a name change is not warranted at all. AI is finally doing some things that CS nerds envisioned in the 70s and 80s. Changing the name when it finally accomplishes what was always envisioned as AI is moving the goalpost, IMO.

FTR, I don't really care all that much. I was just commenting and half joking that a name change is dumb and it never would have happened if it wasn't for Trump simply trying to add something else to his legacy however meaningless and dumb it is.

Mathematicians that know how o use the tools available will excel. It still takes someone to direct the computers.

ntxVol said:

---

JB99 said:

---

ntxVol said:

---

More bold claims by OpenAI. Nothing of significance has been verified or replicated yet.

---

They solved over 700 proofs that were verified. Try to read next time.

---

Anyone who places any value on OpenAI's claims is a fool. Nobody is going to believe them until they actually publish the model and allow mathematicians to replicate the results. If you actually read the article, you would know this.

Quote:

---

But OpenAI's reputation, among mathematicians, for bold claims and little transparency is causing considerable skepticism.

---

---

So it wasn't the LLM that did it, it was some super genius mathematician they hired to do the proofs?!

Logos Stick said:

---

ntxVol said:

---

JB99 said:

---

ntxVol said:

---

More bold claims by OpenAI. Nothing of significance has been verified or replicated yet.

---

They solved over 700 proofs that were verified. Try to read next time.

---

Anyone who places any value on OpenAI's claims is a fool. Nobody is going to believe them until they actually publish the model and allow mathematicians to replicate the results. If you actually read the article, you would know this.

Quote:

---

But OpenAI's reputation, among mathematicians, for bold claims and little transparency is causing considerable skepticism.

---

---

So it wasn't the LLM that did it, it was some super genius mathematician they hired to do the proofs?!

---

IDK, you would have to ask Sam Altman but, don't expect an honest answer.

ntxVol said:

---

Logos Stick said:

---

ntxVol said:

---

JB99 said:

---

ntxVol said:

---

More bold claims by OpenAI. Nothing of significance has been verified or replicated yet.

---

They solved over 700 proofs that were verified. Try to read next time.

---

Anyone who places any value on OpenAI's claims is a fool. Nobody is going to believe them until they actually publish the model and allow mathematicians to replicate the results. If you actually read the article, you would know this.

Quote:

---

But OpenAI's reputation, among mathematicians, for bold claims and little transparency is causing considerable skepticism.

---

---

So it wasn't the LLM that did it, it was some super genius mathematician they hired to do the proofs?!

---

IDK, you would have to ask Sam Altman but, don't expect an honest answer.

---

There are only two answers:

1) the LLM did it, 2) some godlike mathematician came out of nowhere and created the proofs

The proofs are published to github for all to see!

Logos Stick said:

---

ntxVol said:

---

Logos Stick said:

---

ntxVol said:

---

JB99 said:

---

ntxVol said:

---

More bold claims by OpenAI. Nothing of significance has been verified or replicated yet.

---

They solved over 700 proofs that were verified. Try to read next time.

---

Anyone who places any value on OpenAI's claims is a fool. Nobody is going to believe them until they actually publish the model and allow mathematicians to replicate the results. If you actually read the article, you would know this.

Quote:

---

But OpenAI's reputation, among mathematicians, for bold claims and little transparency is causing considerable skepticism.

---

---

So it wasn't the LLM that did it, it was some super genius mathematician they hired to do the proofs?!

---

IDK, you would have to ask Sam Altman but, don't expect an honest answer.

---

There are only two answers:

1) the LLM did it, 2) some godlike mathematician came out of nowhere and created the proofs

The proofs are published to github for all to see!

---

So, if the answer is 1, then at what cost? That is the real question. If it actually cost millions in compute time, then these results aren't very useful. That is where their past claims have fallen short.

YouBet said:

---

TexAgs91 said:

---

YouBet said:

---

I refuse to call it SI. It's AI until it solves scarcity.

---

Why? Actual intelligence hasn't solved scarcity.

Insisting on calling it Artificial Intelligence is like insisting that airplanes perform artificial flight.

---

Exactly.

It's been known as AI since inception of the concept decades ago. It hasn't done anything to warrant a rebranding that suggests it's anything beyond even artificial.

I'm just saying to keep it as is until it solves something practical to warrant being called a Super Intelligence which implies something above and beyond sentience.

---

Quote:

---

GeorgiAg said:

Agree. I don't understand why Trump wants to change the name.


Just another example of wanting to stamp his legacy on something. Anything. It's dumb.

---

Conjecture: TDS is real

Theorem: Your post

YouBet said:

---

I refuse to call it SI. It's AI until it solves scarcity.

---

Listening to conservative AM radio a week or two ago, they were facepalming at calling it SI. They ultimately determined we should call it "augmented intelligence" rather than artificial or super. Even better, it still gets called AI.

LarryLayman said:

---

YouBet said:

---

TexAgs91 said:

---

YouBet said:

---

I refuse to call it SI. It's AI until it solves scarcity.

---

Why? Actual intelligence hasn't solved scarcity.

Insisting on calling it Artificial Intelligence is like insisting that airplanes perform artificial flight.

---

Exactly.

It's been known as AI since inception of the concept decades ago. It hasn't done anything to warrant a rebranding that suggests it's anything beyond even artificial.

I'm just saying to keep it as is until it solves something practical to warrant being called a Super Intelligence which implies something above and beyond sentience.

---

Conjecture: TDS is real

Theorem: Your post

---

lol. I'm one of the last people on here to ever have TDS. I make fun of those people. You must not hang out here much.

This is a simple strike right down the middle to call on Trump.

BboroAg said:

---

GeorgiAg said:

---

YouBet said:

---

I refuse to call it SI. It's AI until it solves scarcity.

---

Agree. I don't understand why Trump wants to change the name.

---

Trump changing the name is a money grab for Trump cronies….how?….all domain name variations of "super intelligence" were purchased before Trump's order

---

Well............

At least that's a reason for the name change that actually 100% makes sense lol

The simplest answer is usually the correct one after all.

Expensive does not mean useless.

A mathematical proof can settle a problem permanently and support future discoveries. The value depends on what foundation it lays, not just the computing bill.

I can guarantee when it solves those equations it isn't using common core math.

ntxVol said:

---

JB99 said:

---

ntxVol said:

---

More bold claims by OpenAI. Nothing of significance has been verified or replicated yet.

---

They solved over 700 proofs that were verified. Try to read next time.

---

Anyone who places any value on OpenAI's claims is a fool. Nobody is going to believe them until they actually publish the model and allow mathematicians to replicate the results. If you actually read the article, you would know this.

Quote:

---

But OpenAI's reputation, among mathematicians, for bold claims and little transparency is causing considerable skepticism.

---

---

Its verified by the actual tool mathematicians use to verify proofs. I dont know what else you need. The skeptics are saying the model had mathematicians work in the training data, which is essentially claiming it stole their work. OpenAI says it did not. But you wouldn't know that since apparently you can't read.

JB99 said:

---

ntxVol said:

---

JB99 said:

---

ntxVol said:

---

More bold claims by OpenAI. Nothing of significance has been verified or replicated yet.

---

They solved over 700 proofs that were verified. Try to read next time.

---

Anyone who places any value on OpenAI's claims is a fool. Nobody is going to believe them until they actually publish the model and allow mathematicians to replicate the results. If you actually read the article, you would know this.

Quote:

---

But OpenAI's reputation, among mathematicians, for bold claims and little transparency is causing considerable skepticism.

---

---

Its verified by the actual tool mathematicians use to verify proofs. I dont know what else you need. The skeptics are saying the model had mathematicians work in the training data, which is essentially claiming it stole their work. OpenAI says it did not. But you wouldn't know that since apparently you can't read.

---

Until the model is published and the results replicated, OpenAI's claims are worthless.

This is weird

BREAKING: President Trump says the White House considers anyone who uses the term “Artificial Intelligence” the enemy. pic.twitter.com/wX9Ln9fcQW

— The Kobeissi Letter (@KobeissiLetter) October 8, 2026