A US CS professor published an article: The mathematical doomsday is coming, humanity's thousand-year-old mathematical palace has been destroyed by AI.

OpenAI has sparked widespread outrage among mathematicians across the globe!

On October 6, it released 722 papers on major unsolved mathematical problems all at once.

Scott Aaronson, a US computer science professor and ACM Fellow, exclaimed: The mathematical doomsday is coming!

Mathematicians worldwide have united to launch a collective boycott against OpenAI.

Terence Tao released a joint statement led by AMH on his personal blog, openly breaking ties with OpenAI.

In the statement, they refer to OpenAI as follows: "Yesterday, OpenAI — the company currently embroiled in lawsuits over allegations of illegal plagiarism, copyright infringement, and trademark dilution — released a batch of manuscripts..."

They describe OpenAI's actions as "naked computing power hegemony".

Yann LeCun, the Turing Award laureate, did not side with mathematicians this time, and his view is: We have discovered a new continent.

The Mathematical Doomsday Arrives: "Mom, You Got Outperformed by a Machine in Seconds!"

Recently, Scott Aaronson, a renowned American theoretical computer scientist and professor at the University of Texas at Austin, published a long essay titled *The Mathematical Doomsday*.

The essay opens with this poignant scene.

"Mom, I heard you got instantly outperformed! I heard a robot solved a math problem you have been studying your entire career! Oh no!"

His 9-year-old son was taunting his own mother, the renowned complexity theorist Dana Moshkovitz. The innocent words of a child hit hard with every syllable.

Among the hundreds of proofs released by OpenAI lies the complete proof of the Unique Games Conjecture (UGC) by Subhash Khot.

This is a conjecture that has plagued the academic community for decades, and has been the target Dana Moshkovitz has spent her entire career tackling. Now it is easily solved by an AI.

Dana is by no means the only "victim". In this paper tsunami that resembles a "fire sale", the sky has fallen for computer scientists and mathematicians.

Many mathematicians woke up, looked at the AI-generated files on their screens, and nearly fainted on the spot.

Aaronson directly named this day "the doomsday of the mathematics community".

Any randomly selected achievement is enough to win the "Best of the Year" award in its field, and even cause a huge stir.

L=BPL Derandomization: It proves that probabilistic logarithmic space and deterministic logarithmic space are the same thing, directly solving a long-standing unsolved case in the theoretical computer science community.

New Bound for Matrix Multiplication: O(n^(9/4)) time complexity! It breaks the deadlock that has stagnated for years, and uses a completely new method that no human has ever conceived.

Fourier Transform Breaking the 1960s Barrier: The time complexity of integer multiplication is reduced to below O(n log n).

What is most chilling is the "cost" of these achievements.

The problems are solved only by the latest internal model (at the GPT-Pro level) that is about to be launched to paid users. For each top-tier difficult problem, the AI only spends about 3 hours of computing power on average.

In the time it takes a human to drink two cups of coffee, the AI has solved 5% of the most difficult problems in the human top-tier question bank.

Unintelligible "Alien Creation": Humans Reduced to Porters on the Ruins

The deeper despair is that what the AI writes is completely unintelligible to humans.

After going through the proof of the UGC conjecture, Dana fell into deep confusion:

It feels like something written by a person on hallucinogens. There are too many unclear and illogical parts.

It invented a completely new, extremely strange encoding method with noise tests. It is not a long code, nor a short code — it is some kind of alien-like crazy thing.

The paper is written terribly, with chaotic citations and full of jumps that humans cannot understand.

In the end, human mathematicians have to turn on another AI (such as Astra) to help them "translate" and sort out this paper written by AI.

This is the most absurd scene brought by the "mathematical doomsday": Human scholars are no longer knowledge creators, but have been reduced to "archaeologists" and "translators" in the ruins of alien civilization.

Jordana Cepelewicz, a writer for *Quanta Magazine*, came up with a brilliant metaphor.

It's as if you were instantly teleported to the top of a high mountain.

Mist surrounds you, and you have no idea where you are. You have never experienced the hypoxia of climbing, nor have you developed the tools to scale cliffs.

You are just sitting on the dark top of the mountain, while the maker of the teleportation machine coldly tells you: It knows this mountain better than anyone else.

Yes, the truth is right there, but humans have lost the process of climbing.

When we solve problems in a non-human way, what on earth does it mean to truly "know" a thing?

Mathematicians Are Furious: No One Asked You to Solve These Problems!

Why did human mathematicians stand on the opposite side of OpenAI overnight?

Because OpenAI's actions are so extreme that it directly tramples on the dignity of top human brains.

Some time ago, OpenAI once publicly claimed that it had solved a Millennium Prize problem. At that time, Terence Tao already sensed the danger.

He immediately joined 25 Fields Medalists to issue a joint appeal, hoping that commercial companies would slow down:

This is absolutely crazy! AI companies are accelerating endlessly, knowing nothing about what will happen afterwards. We must slow down. There is no reason to go so fast at all, not a single reason!

But what was the result? OpenAI didn't take the warnings of these Fields Medalists seriously at all, and immediately released 722 machine proof files, slapping the entire mathematics community in the face with the most violent means.

The Association of Human Mathematicians (AHM) is the "industry defense line" spontaneously established by top mathematicians around the world. The core purpose of AHM is only one: to guard human understanding of the essence of mathematics, and to prevent academic norms from being completely destroyed by unrestrained violent computing power.

In this joint statement led by AHM and reposted by Terence Tao, mathematicians hit back: "No one asked you to do this work! No one ever asked you to solve these proofs!"

"Releasing more than 700 files at one time is not scientific research at all, it is purely showing off computing power hegemony! We strongly urge all mathematicians to stop cooperating with OpenAI and return to the scientific vision centered on human understanding."

Some people support it, while others mock: If AI conquers hundreds of terminal illnesses, the medical community will definitely welcome it warmly.

Turing Award Winner Yann LeCun: On the Contrary, We Have Discovered a New Continent!

Professor Florent Krzakala from EPFL in Switzerland said bluntly: Some parts of mathematics are already dead.

However, Turing Award winner Yann LeCun put forward a completely different assertion.

He said: On the contrary, mathematics is not dead at all, but is ushering in an unprecedented new era!

In LeCun's view, mathematicians do not need to be so panicked. "Formal proofs will be greatly automated, but that does not mean humans lose their value. The focus of future mathematical research will shift to developing new concepts, new abstractions, new definitions and new conjectures."

Then he gave this metaphor: "The invention of the ship did reduce the importance of humans 'swimming' by themselves. But the ship allowed us to discover new continents!"

In the comment section of Terence Tao's blog, this kind of division can also be seen.

A scholar signed Jörg Neunhäuserer confessed that after the three problems he had spent a lot of time studying were solved by AI, he did not feel deprived of anything, but instead "felt a huge sense of relief".

Because he no longer has to do these drudgery, and the machine has proved that his intuition is correct.

A doctoral student majoring in algebraic geometry also expressed opposition to this view.

The Horrifying Hidden Shadow: Why Is Cryptography the Only Missing Field?

In this mathematical doomsday, there is another creepy doubt.

The hundreds of papers released by OpenAI cover almost all fields including number theory, combinatorial mathematics, and quantum complexity.

But cryptography is the only field that is missing.

Why? Did AI hit a wall in cryptography?

Scott Aaronson revealed a chilling inside message: All AI companies have now begun to investigate whether their latest internal models can crack the most important cryptographic protocols in the world today.

If AI can decrypt top-level complex theoretical architectures and invent "alien codes" that even top human brains cannot understand, then are human bank encryption, network security, and even national-level secrets (RSA/AES, etc.) just a thin layer of paper in front of it?

If they have already been cracked, "rush to make preparations before the rest of the world finds out" — this is what the major giants are doing behind closed doors in secret right now.

In the long essay by Professor Scott Aaronson, he wrote —

Then they went after Navier-Stokes, and I said nothing, because I never studied Navier-Stokes. But when they went after RL and L, I realized that things were getting serious.

Who knows that the next target won't be their own research field?

On that shocking night, Scott Aaronson did not continue to pore over those hundreds of alien papers. He decided to turn off his computer and watch a movie with his children.

He thought this movie could provide the most practical survival guidance for his children.

The name of this movie is *Terminator 2*.