The Unsympathetic Nature of Mathematicians

The upshot of the whole "AI Solving Mathematics" thing is that these people with generally sympathetic viewpoints on their field are suddenly becoming protectionists at a rapid rate while simultaneously struggling to not sound that way. This tension causes them to take positions that are mutually contradictory and do not stand on their own. A typical Hacker News comment reads not differently from Terence Tao's blog or those of many mathematicians on Twitter or Bluesky so it's a good starting point to look at the feelings at play here:

YHacker News

Comment on There's never been a better time to go into math (by: 3blue1brown) [video]

I'd like to believe Grant but I find this assault on mathematics to be dehumanizing and demoralizing. AI has destroyed a major source of "life's meaning" for many people who pursued mathematics. Imagine waking up one morning to find that an alien technology has solved the problem you had dedicated your life to (and many more to boot.) Not another mathematician but an alien technology. As a scientist, I don't want to end up as a laboratory assistant to an AI. I can take not being the smartest human in the room, but I'm pretty sure I don't want to spend my life serving a silicon master.

nritchie•2026-10-18

The Arguments

The Illegibility Argument

One argument for why computer proofs are rubbish is that computer proofs are often illegible or extremely forced and consequently they do not invite new lines of reasoning, new fields of mathematics, or new angles of attack on other problems. OpenAI's proofsets are, in effect, a binary blob thrown upon the masses for which no source code was ever written - the compiler having written a compiled program to start with. But if this is true, and if the value in Mathematics has ever been the creation of the legible proof then surely one cannot have dehumanized Mathematics. In fact, one hasn't anythinged Mathematics and simply provided binary true/false statements of not much utility.

If this were the case, then human mathematicians still carry the predominant amount of value, and AI proofs are a mere curiosity that happen to exist. It's hard to argue then that OpenAI's release has caused any damage. At best, they have proven themselves useless.

Incentive Structure

Some claim refuge in complaints that the real damage that's done here is to policy.

YHacker News

Comment on "I'm on a war footing": Terry Tao and the crisis in math

Yes there's still useful work only mathematicians can do, what is the incentive system that encourages them to do it? Currently, none.

pebbly_bread•2026-10-19

But only some 43% of mathematics publications are federally funded, and that scheme of funding is less than a hundred years old. And certainly it's hard to claim that the dehumanization of mathematics has been caused by a lack of a highly-competitive grant-writing and reviewing system. But perhaps it is. Perhaps the beauty of mathematics is in receiving research grant money.

The Self-Actualization Argument

This one is particularly galling. Supposedly, it is not enough for truths to be discovered, but also that they be done so in a process that makes specific practitioners of specific occupations feel some sort of exalted sense of self. None of this is believable. And for a group that is predisposed to make particularly logical arguments in general, the general scatteredness of reasoning illustrates that something is stopping them from realizing the wonder of what we're going through.

The system myth that we are taught about our common form of taxation in government is that our forced contribution into a collective is a mechanism for us to provide society with utility in a manner commensurate with the utility that it is providing us so that a virtuous loop of utility can be maintained.

If the entire purpose of Mathematics grants was simply to provide a few people with personal utility, that raises the question of whether we could not serve some greater good by using AI to get the results and then simply serving high-purity fentanyl to some larger number of humans. This way we would get results faster and provide personal joy to a larger group of people.

The Cheat-Code Argument

A variant of The Whispering Earring, this argument claims that being informed of the answer removes the ability to form a contributory theory of why.

YHacker News

Comment on "I'm on a war footing": Terry Tao and the crisis in math

> That’s what technology does. And he might be upset but we’re the ones paying for it and frankly, I’m happier to have more questions answered than for fewer answered but in an artisanal fashion.

Would you prefer if graph theory didn't exist but we could answer if you can pass all bridges in Königsberg, without going through same bridge twice?

Or if we could answer any question without knowing why? What about answering Hilbert without theory of Godel's Incompletness theorem? Or knowing Nth digit of Pi without calculus.

Because that's essentially what you're asking for. If you gave humans a Djinn that
could answer perfectly any question, it would kill most of our endeavors. Because it's not a problem that was unsolved for years "It's just the thing Djinn already answered so why bother?" Even if one is extremely curious, and devote a chunk of their life answering it, others would not see value in that curiosity. Because a Djinn can answer it in a few minutes.

Ygg2•2026-10-18

But this faces a bigger problem: if discoveries preclude other work into them, then the biggest danger is releasing mathematical results without first proving that they are the apotheosis of mathematical results in a field. What theories have we been blinded to by the creation of graph theory? We might never know. The human-djinn who gave us that must necessarily have killed off other branches through providing this particular theorization. Any mathematical result now murders countless other endeavours. Only results provably encompassing all they destroy could reasonably be permitted.

Complying with this would require book burning and the rampant mathematicide of those who are aware of our present theories until we can build a system that only releases mathematical constructs that are provably destroying the least amount of other theories.

Now personally, I find this conclusion absurd, and I also find it absurd that being told that many $2^n-1$ are prime would make it impossible to explore why.

The Ramanujan Problem

Another problem with the djinn is that he existed, and he is not considered the enemy of Mathematics. Providing factual expressions and results without explanation, Ramanujan nonetheless is considered a hero of sorts. Perhaps that title should be given to Hardy, and good old Srini should be relegated to the dustbin of history - a mere machine providing "treated waste". I think not, and I think his existence and his status indicates that this entire argument is bullshit.

Curiosities

The Fragility of Human Will

A popular piece of discourse about The Three Body Problem is that it is unbelievable that scientists would react to meaningless results with despair. Many believed that each failed experiment would offer insight into new ways the universe could operate. The physicists would embrace these failures with enthusiasm. Or so we were told. But this example of the Mathematicians proves the fragility of human endurance. A man may have his leg chopped off and yet volunteer for further battle, but his fellow man, informed that the thing he has been seeking has been found will hate the discoverer.

Perhaps physical damage is endurable but not being beaten to the punch. The Sophon should not have given us falsehoods. Our finest minds would apparently have been turned to despair instead by fact. "Yes, it's Dark Matter in the Bullet Cluster" says the Sophon, and the Astrophysicist is rendered mindless, striding off into the sea an unrequested martyr.

The Mischaracterization of Desire

Another mistake is a mischaracterization of what makes oneself tick, typified in the sort of self-claim made here:

Zach Weinersmith @ZachWeiner

I think a lot of the psychological differences in reaction to AI automation have to do with to what extent a category of people do their work for a goal vs. for a lifestyle. A lot of us artists got into this "business" because we wanted to spend our time creating things.

Oct 19, 2026

I think this tweet is quite close to accurate about 'the other side'. There are many people who do what they do because they have a goal. One of the things I dearly desire is to exalt intelligence. I wrote programs because I wanted to minimize the amount of wasted effort that Man expends; I wrote blogposts so that if I have some insight or some information that is useful some other intelligence could use it to better itself; and I have always enjoyed seeing people succeed at things that expand this. That is fair enough.

It is not false that I did not write a program primarily because the act of writing a program brought me pleasure. It did, and I have spent a fair bit of my life on my abstract classes and my interfaces, and this and that. But ultimately, the thing I desired was the outcome.

But I think the other thing is not true. It is not actually true that Mathematicians produce Mathematics so that they may have discovered some Mathematics. In fact, were a Mathematician to spend his time in a cave, find some remarkable proof and leave the cave to find that someone else had discovered it less than a day earlier his reaction would not be the continued pleasure of having discovered the proof. In fact, the sentiment he will likely express is disappointment for having been scooped. The primary motivation for the proof discovery is not the discovery then, but because discovery is rivalrous in Mathematics. The one who discovers is The One Who Discovers and the other who independently discovers is No One.

This is not a pedantic mischaracterization. It is fairly fundamental. Whereas many people draw an immense amount of pleasure from trying to play a riff that Steve Vai came up with, it seems disastrous to a mathematician to perform merely a copy of a similar master's work. The Intellectual Property is critical. Writers prosper from translated renditions of famous works, even today. Not so mathematicians. Inscrutable works valueless without comprehensibility yet supposedly render valueless that which provides insight? Unconvincing.

Conclusion

It is true that the most intelligent people can construct the most elaborate rationalizations to protect their unfounded beliefs. And as Sinclair's Maxim predicts, many have arrived at these rationalizations. The truth appears closer to Gavin Belson's famous line:

I don't know about you people but I don't want to live in a world where someone else makes the world a better place, better than we do

— Gavin Belson, Silicon Valley