Top
Best
New

Posted by ipnon 1 day ago

What's a mathematician to do? (2010)(mathoverflow.net)
166 points | 80 comments
getnormality 1 day ago|
From one of the answers:

> mathematics only exists in a living community of mathematicians that spreads understanding and breaths life into ideas both old and new. The real satisfaction from mathematics is in learning from others and sharing with others. All of us have clear understanding of a few things and murky concepts of many more. There is no way to run out of ideas in need of clarification.

Yes! And this applies to all human culture, not just math. Everything people have figured out needs to be in living form to carried on. The more people the better. If math, or any product of human skill, is only recorded in papers or videos, that isn't the same as having millions of people understanding it in their own ways.

Modern culture often emphasizes innovation and fails to value mere maintenance, tradition, and upkeep. This can lead to people like the OP feeling that they have nothing to contribute, when actually, just learning math, being able to do it, being able to help others learn it - all of these are contributions.

We are all needed to keep civilization afloat, in ways we cannot anticipate. We all need to pursue some kind of excellence just to keep human culture alive.

ajs1998 21 hours ago||
This is why I think Brady Haran is one of the coolest living mathematicians. Numberphile is educating a new generation of young mathematicians for anyone with access to youtube. Accessible math communication is so important. So many cool things are buried in textbooks and papers the average person would never read.
thaumasiotes 20 hours ago||
Numberphile doesn't do any education. That's like saying the Discovery Channel is educating a new generation of zoologists.
ajs1998 18 hours ago|||
I simply don't care to gatekeep what counts as education. It has taught me things from videos I can still recall a decade later and pushed me to explore different areas of math I wouldn't have done otherwise.

It's education for whoever finds it educational

srean 5 hours ago||||
"If you want to build a ship, don’t drum up the men to gather wood, divide the work, and give orders. Instead, teach them to yearn for the vast and endless sea" -- Saint-Exupery

This is exactly what Numberphile does. Those who are hooked will find the resources on their own. They need a reason to look for them and Numberphile gives them one.

whatever1 15 hours ago||||
I believe that I have learned cumulatively double in the past 10 years from YouTube compared to what I learned in 6 years of middle and high school. And I don’t spend 8 hours per day on YouTube.

Plumbing, react, combinatorics, real analysis, python, c++, cad, micro and macro economics, reinforcement learning, to name just a few of the things I learned through YouTube.

We don’t give enough credit to what we take for granted today.

stevenhuang 18 hours ago|||
If topics are presented engagingly these efforts will no doubt inspire the next generation.

Young people are curious, sometimes all they need is a spark and to be introduced to a new topic in an engaging way. These forms of content deliver that spark.

rdevilla 23 hours ago|||
> Everything people have figured out needs to be in living form to carried on.

It would appear that LLMs are invalidating this claim. Things can live in synthetic form and carry on just fine. Instead of cultivating a population of learned minds we are just feeding a few dozen egregores of models and training corpuses.

pegasus 17 hours ago|||
They are not invalidating this claim, and cannot, unless we'd actually try it out for a few generations. Which we shouldn't and won't.

LLMs are quite good at simulating life and living intelligence (in the short term), but they aren't any of that. That's why we call it artificial intelligence. It's true that we can't put our finger on what exactly the difference is, but it's not like reality has ever felt encumbered by our limited understanding.

gosub100 18 hours ago|||
All LLMs do is launder other people's IP. So I don't think you invalidated any other claim.
susam 19 hours ago|||
> From one of the answers [...]

Thank you for highlighting that answer. It is one of my favourite pieces of writing about the culture of mathematics. I just want to add that that particular answer is now affectionately known as Thurston's Paean.

DrewADesign 23 hours ago|||
In theory, sure. In practice, our society is a) not set up to value things which don’t have an immediate financial ROI, b) is valuing them less as time progresses, not more, and c) is experiencing some very serious transitions that may destroy the financial viability of devoting a lot of your time and energy to some very important things.
netcan 23 hours ago|||
Living culture is a concept that I think is quite unintuitive to modern minds. Examples of it are all around us... but it's usually blatantly missing from our "big picture" thinking.

For example. Take a modern country with a modern economy. Flatten it. Destroy all the factories. Bankrupt all the companies. You can get back to a fully modern economy again quite quickly. WWII demonstrates it.

Taking an unindustrialized country through the development process... that's very tricky. It can't really be rushed.

For a long time, economic development was seen as mostly capital and technology. You need time to develop all the capital needed. Roads, factories, etc. But... development efforts underperformed. Then the idea of "human capital" got popular as a way of explaining the deficit. Education, mostly. Development efforts still underperformed.

I think the "living community" thing is the answer to this. It' ecology. You can't make a rainforest by just dumping all the necessary organisms into the right climate. It's the endlessly complicated relationships between all those organisms that make the rainforest.

This is one of the things that worries me about the pace of modern change. When writing and literacy resurged in classical antiquity... we totally lost all the ways of (for example) doing scholarship orally. Socrates (through Plato) wrote about some of the downsides to this.

...and we did completely lose oral scholarship. We have no idea how to do it. Once the living culture died... it stayed dead. All the knowledge contained within it went away.

sendes 15 hours ago||
> I think the "living community" thing is the answer to this. It' ecology.

I agree. A body of knowledge, mathematical knowledge being one of them, is a give-and-take between its producers and consumers; a market for ideas. It grows in that ecology of people with its pathfinders, specialists, generalists, historians, educators, etc. Committing to a body of knowledge is becoming part of its living culture.

Where I disagree: I believe some of the loss is inevitable. Keeping in mind the example of a body of knowledge above, as the scale of what's accumulated until now grows, the role each of us play in the sustenance of its culture shrink. This is a direct consequence of the modern developmental process (ie division of labour to the point where it feels like we are all modular parts of a much larger whole).

I can't say whether its better to focus on recovering what's lost, or, trust the process, as it were.

whattheheckheck 16 hours ago||
But were trapped in only keeping alive that has a reward signal which is can it help pay rent, pay for food, get love, etc

Theres a limited amount of time, space and energy so what's the ideal mechanism to say what to pay attention to or not

bloaf 23 hours ago||
So I've got a gut feeling that math (like human languages (like programming languages)) is best learned in service of some greater end.

I look at some truly impressive projects like CLASP which sprang into existence not because of someone noodling around, but because they had a bigger goal which required the team build it.

So my advice to any mathematician who feels lost, like they don't know what to work on, would be to go collaborate with someone who has an actual goal, to look for inspiration in the kinds of math they need.

Today, there are a lot of opportunities to jump forward that only get capitalized on through coincidence (e.g. two people bump into each other at a conference, or researcher happens to have a colleague working on a related problem through the lens of a different discipline). If AI does nothing but guarantee that everyone will have such a coincidence by serving as that expert from a different discipline, that will still be a massive driving force for progress.

The question of "whats a mathematician to do" is still clear: you need to find and curate and clearly express interesting and valuable problems.

nimonian 23 hours ago||
It's a delightful counterintuition that your gut feeling is mostly wrong: https://webhomes.maths.ed.ac.uk/~v1ranick/papers/wigner.pdf

Far from being motivated by some applications, the most useful discoveries in mathematics are usually discovered "for their own sake" and their application is only discovered later. Sometimes centuries later!

skybrian 22 hours ago|||
If so that seems like an opportunity for people who want to work on applied math? There’s a big backlog of techniques that so far have not been useful.
kcexn 13 hours ago||
Absolutely! The backlog is enormous though, and much of mathematics requires a great deal of work to understand it to the depth required before a novel application becomes apparent.
transitorykris 22 hours ago||||
Parents reads as a comment on the usefulness of applying mathematics to problems in the world (applied mathematics) and discovering mathematical problems that push mathematics forward (pure mathematics) in the process. Pure mathematics is incredibly important, but I’d hardly count it as useful if we need to wait centuries.
aleph_minus_one 14 hours ago||
> but I’d hardly count it as useful if we need to wait centuries.

This is not the fault of the mathematicians.

dennis_jeeves2 22 hours ago|||
>are usually discovered "for their own sake"

Like prime numbers? (used in cryptography)

jvans 23 hours ago||
Lots of fun counter examples to this. Complex numbers were introduced in the 1600s with no practical application for almost 300 years until they were used in electromagnetism and quantum mechanics.
lmm 10 hours ago||
> Complex numbers were introduced in the 1600s with no practical application for almost 300 years.

On the contrary, complex numbers were introduced to make the cubic formula work.

lisper 21 hours ago||
> One can rewrite their books in modern language and notation or guide others to learn it too but I never believed this was the significant part of a mathematician work

There's yer problem right there. Good pedagogy is hard and highly undervalued. IMHO Grant Sanderson (a.k.a. 3blue1brown) is making some of the most significant contributions to math in all of human history by making very complex topics accessible to ordinary mortals. In so doing he addresses one of the most significant problems facing humankind: the growing gap between the technologically savvy and everyone else. That gap is the underlying cause of some very serious problems.

TrackerFF 20 hours ago||
Big fan of him - but I also want to throw out the most obvious name in this space: Sal Khan

Hard to imagine now, but back when he started out, there were really no (to very few!) accessible math tutoring vids on the video platforms. Most of the times you had some universities, like MIT, putting out long-form vids from lectures - but actually having easily digestible 5 min vids like those Khan put out, just wasn't a thing.

bobajeff 20 hours ago|||
I like to watch 3blue1brown too but I think it's a bit of an exaggeration to say his topics are accessible to normal folks. From my perspective I think it's more realistic to say he makes videos that shows you the beauty in math without having to understand it really. Which is valuable since most people get turned off on math because tiresome drills and tests hammered into them at school by people with zero interest in it.
xiaoyu2006 18 hours ago||
Quite true. Real math needs practice and calculation to build the intuition and motivation towards the abstraction we want to construct. The video is more of a complementary material to the boring lectures (my prof uses 3b1b videos sometimes).
susam 20 hours ago|||
Good pedagogy is a problem even for graduate-level mathematics students and professional mathematicians. The proofs in many graduate-level mathematics textbooks are, in my humble opinion, not really proofs at all. They are closer to high-level outlines of proofs. The authors simply do not show their work. The student then has to put in an extraordinary amount of effort to understand and justify each line. Sometimes a 10-line argument in a textbook might expand into a 10-page proof if the student really wants to convince themselves that the argument works.

I am not a mathematician, but out of personal interest, I have worked with professional mathematicians in the past to help refine notes that explain certain intermediate steps in textbooks (for example, Galois Theory, by Stewart, in a specific case). I was surprised to find that it was not just me who found the intermediate steps of certain proofs obscure. Even professional mathematicians who had studied the subject for much of their lives found them obscure. It took us two days of working together to untangle a complicated argument and present it in a way that satisfied three properties: (1) correctness, (2) completeness, and (3) accessibility to a reasonably motivated student.

And I don't mean that the books merely omit basic results from elementary topics like group theory or field theory, which students typically learn in their undergraduate courses. Even if we take all the elementary results from undergraduate courses for granted, the proofs presented in graduate-level textbooks are often nowhere near a complete explanation of why the arguments work. They are high-level outlines at best. I find this hugely problematic, especially because students often learn a topic under difficult deadlines. If the exposition does not include sufficient detail, some students might never learn exactly why the proof works, because not everyone has the time to work out a 10-page proof for every 10 lines in the book.

Many good universities provide accompanying notes that expand the difficult arguments by giving rigorous proofs and adding commentary to aid intuition. I think that is a great practice. I have studied several graduate-level textbooks in the last few years and while these textbooks are a boon to the world, because textbooks that expose the subject are better than no textbooks at all, I am also disappointed by how inaccessible such material often is. If I had unlimited time, I would write accompaniments to those textbooks that provide a detailed exposition of all the arguments. But of course, I don't have unlimited time. Even so, I am thinking of at least making a start by writing accompaniment notes for some topics whose exposition quality I feel strongly about, such as s-arc transitivity of graphs, field extensions and so on.

zozbot234 19 hours ago||
These days it's easy to just look for the details to any proof on mathlib. Of course a computer checked proof is not always super intuitive for a human, but most of the time it does work quite well.
LPisGood 21 hours ago|||
Indeed, pedagogy is important to staving off the end of mathematics.

That sounds dramatic, but it’s really obvious if you think about it. Right now, a person has to study for about 20 years (on average) to make novel contributions in mathematics. They have to learn what’s come before, the techniques, the results, etc. If mathematics continues, eventually it could take 25 years, or 30 years, or even a whole lifetime. At some point, most people will not be able to understand the work that’s been done in any subfield (or the work required to understand a subfield) in a human’s life. I claim this is the logical end of mathematics, at least as a human endeavor.

Now, there will be some results which refine other work and simplify results, but being able to teach a rapidly growing body of literature efficiently will be important to stave off the end of mathematics.

pertique 21 hours ago|||
There's a Scott Alexandar story that plays with this exact topic: Ars Longa, Vita Brevis [1]

To your point, I think you're right. I'm not in mathematica, but the value of good pedagogy on shrinking the time it takes to get people to the forefront of any field feels like it's heavily overlooked.

https://slatestarcodex.com/2017/11/09/ars-longa-vita-brevis/

rustymather 19 hours ago||
Seconded. Good pedagogy is like fertilizing the soil- it creates conditions conducive to learning in order to do good research.
GuB-42 22 hours ago||
I think it is like for a programmer to ask "How can one contribute to computer science?", while thinking about people like, Dijkstra, Knuth, or maybe even Carmack.

There are some geniuses who do groundbreaking work, but this wouldn't be of much use it it wasn't for the millions of people who do actual work with these theories (applied math), and teachers who train the next generation. In the academia, small discoveries exist too, these can be the stepping stones for the big things to come, even if they don't have a direct application now.

Lerc 21 hours ago||
>I think it is like for a programmer to ask "How can one contribute to computer science?", while thinking about people like, Dijkstra, Knuth, or maybe even Carmack.

I had a conversation a day ago with a couple of high-school students who, were obviously smart (and a bit on the spectrum), but also lacking in broad knowledge, as one would expect from high-school students.

I think it revealed something about that sense of 'magic' or inaccessible talent. One of them mentioned the fast inverse square root function and marvelled about how even anyone could even come up with an idea like that because it seemed to him to be some transcendent feat to be able to realise casting binary floats into ints would be useful. They had looked at the function and couldn't fathom how something like that worked.

We had no computer to hand, but I asked him what 10^100 divided by 10^10 was. Smart kid that he was, answered instantly, correctly. I then asked him what operation he performed in his head to do that, and noted floating point exponents are just using 2 instead of 10. On the spot he figured out how the fast inverse square root worked. The magic vanished and it became accessible. Some people lament the loss of 'magic' in this way, but I think the thing that makes it special, in this instance and in the universe in general, is that it still works, and it isn't magic. It's real and the fact that it be that without invoking some unaccessible property makes it even more special

bee_rider 22 hours ago|||
I do think this has been less pressing of a question for programmers. For the longest time there was infinite work to do, no matter your depth—implementing business logic, making frameworks more general, making sure things fit into cache lines—that was accessible to us non-geniuses. Maybe in the LLM era this sort of t-crossing and i-dotting will go away.
zozbot234 21 hours ago||
> Maybe in the LLM era this sort of t-crossing and i-dotting will go away.

On the contrary, prompting LLMs creates a whole lot of newly accessible basic work.

s5300 22 hours ago||
[dead]
beyonddream 15 hours ago||
I am glad to see this today - after reading Tim gower's recent post on chatgpt 5.5 pro's phd level research ability I was feeling slightly sad about the future of math research.

Interestingly enough, the moment I saw the title I thought of Bill Thurston's famous article "On proof and progress in mathematics" and the top comment on the OP's thread is from him! Reading his reply sort of gave me the antidote to the temporary blues I felt yesterday.

fasterik 14 hours ago|
Why did it make you sad? I can see why it would make training researchers harder, but once someone attains a postdoc-level skill in mathematics research, wouldn't having a PhD-level AI assistant just boost one's ability to do more ambitious research?
beyonddream 12 hours ago||
You should read that essay! The model is capable of producing phd level research on it's own with minimal set of prompt's. I was sad because of this paragraph:

"That view is that there is still a great deal of value in struggling with a mathematics problem, but that the era where you could enjoy the thrill of having your name forever associated with a particular theorem or definition may well be close to its end. So if your aim in doing mathematics is to achieve some kind of immortality, so to speak, then you should understand that that won’t necessarily be possible for much longer — not just for you, but for anybody."

He may seem to imply the end is only for some subset of reasons but if you read the entire essay he is just trying to give hope where the rest of the essay is really damning!

Schlagbohrer 1 day ago||
After reading another post about the most recent advances LLMs have made in finding and writing up novel, correct proofs, it sounds like the frontier models are now at the point of PhD student level. I wonder how a math student could contribute today, if they're just starting on the PhD track? Maybe by using LLMs as a mighty tool and providing skilled usage and oversight?

It must feel similar to those who wanted to become chess or go masters after computers surpassed humanity in those games.

zozbot234 21 hours ago||
> After reading another post about the most recent advances LLMs have made in finding and writing up novel, correct proofs, it sounds like the frontier models are now at the point of PhD student level.

This is somewhat misleading, the LLMs' contributions are in a limited niche of highly technical problem solving. They're neat but they're not the first mathematical theorem that gets automatically solved by a computer, that was done already in the 1990s.

> Maybe by using LLMs as a mighty tool and providing skilled usage and oversight?

Yes, even in the areas where LLMs are at their best, we'll still need a lot of human effort to make the results cleanly understandable. LLMs cannot do this well, even their generated papers have to be rewritten by human experts to surface the important bits.

DroneBetter 17 hours ago||
> done already in the 1990s by human-written programs that iterated through the finite casework that human thought had reduced the theorems to (four-colour theorem, FLT, etc.), which recent developments (eg. LLMs autonomously resolving Erdős problems) seem meaningfully distinct from. > human effort to make the results cleanly understandable well, perhaps loops of "derive proof through reasoning in English, formalise in Lean, use AST size of formal proof as a metric to optimise (via an LLM-guided search), translate back into English" could improve this? a lot of resources are being spent to make frontier LLMs more resistant to hallucinations via Lean, perhaps cogency will increase as a byproduct.
dataviz1000 23 hours ago|||
LLM models can only predict the next token.

The can't predict the consequences of an action predicting one token after another. They can't solve a Rubik's Cube unlike a 7 year old human who can learn to do it in a weekend. They can't imagine the perspective of being a human being unlike a 7 year old human if asked to imagine they where in the position of another human.

DoctorOetker 23 hours ago||
Those are very strong claims, do you really believe an LLM can't be trained to solve Rubik's Cubes?

Can you imagine what if feels like to be a LLM?

Can one LLM have a better sensation of what it feels like to be a different LLM (say one that score a little better?)?

You design circularly defined criteria...

r_lee 23 hours ago||
honestly I'm pretty sure opus could solve a rubiks cube if you just gave it the layout if the sides and looped until it would solve it

or even just take a picture of the thing, since they can digest visual input now

generic92034 1 day ago|||
I wonder if AI is one means to overcome the natural limits of human knowledge aggregation [0].

On the other hand, in the very long run, what does it mean if a talented human being does not have enough years of life to fully analyze and understand an extremely advanced proof created by AI?

[0]: https://slatestarcodex.com/2017/11/09/ars-longa-vita-brevis/

morkalork 23 hours ago||
Perhaps it will become like those cathedrals that took centuries and many generations of humans to build.
generic92034 22 hours ago|||
Yes, but you (as a human) can still understand the cathedral (the building). This is not guaranteed for advanced AI work in mathematics in the future. If so, are we/they are really still adding to human knowledge, at this stage?
TimorousBestie 23 hours ago|||
Mathematics as an aggregate already is that cathedral. It is grander and more beautiful than any earthly cathedral.
photochemsyn 22 hours ago|||
If your motivation is being recognized as the best of the best, winning the competition, yes it’s probably a bleak world. But if you motivation is improving your own capabilities, with the metric being if you’re better know then you were last month, then it’s not a bleak world, there are many more tools available to help you learn and improve now then there were in the past.
4qshT 23 hours ago||
The Mathoverflow question was asked 15 years ago. The top answer says that the human community part is very important and spreading knowledge an critical thinking is valuable.

The most recent advances are stunts by a handful of famous prompters who are funded in various ways by the LLM industrial complex.

How many theorems are proven by mathematicians each year? Let's guess 10000. Then the Erdos toy proofs with unknown token and resource usage are less than 1%.

sdenton4 22 hours ago||
...And in 1900, how many carriages were horseless?
ppll_12 22 hours ago||
In 2026, how many people X-ray their feet at the shoe store or have watches with radium paint?

Ironically, there is a shoe company pivoting to AI. My taxi driver told me buy the stock:

https://www.bbc.com/news/articles/c98mrepzgj7o

ashley95 16 hours ago||
I think we also have to be honest and admit that, yes, indeed, there is less novel maths for all of us to be doing. The pioneers came first and discovered a lot of low hanging fruit. There were a lot of geniuses that mined the rest and reached higher in the tree. Now even the smartest mathematicians are left solving abstract puzzles with little utility in the real world. (Don't get me wrong, it's very fun, and sometimes useful too.)

After my PhD in applied mathematics, I decided to leave the field, partly because I feel it really has advanced so far that new discoveries do little to move the needle in the real world. There's enough smart people who obsess over nothing else but maths that I can go and do more practical stuff...

_alternator_ 16 hours ago|
Is this not just a perspective issue? The fields we learn about in school are the ones that are in some sense mature, that have been cashed.

When I was in grad school, we learned about wavelets, but we did research on convex optimization for statistics. The first was an accomplishment of the last generation of mathematicians, and it would be hard to publish something groundbreaking. But nobody had really considered sparsity inducing optimization, so that was our problem.

In many ways, the situation is somewhat better for applied mathematicians because the problem space is wider. Ingrid Daubechies was an applied mathematician, and her work on wavelets was an outgrowth of work that originated in the petroleum exploration community. Wild how these connections get made.

jebarker 23 hours ago||
“Comparison is the thief of joy.”

Do the math because you enjoy doing the math and if you do it long enough you may well do something of value to someone else. Same goes for most intellectual and artistic pursuits I think.

I’ve learned for myself that as soon as enjoyment is based on some future achievement or ranking my work against others the day to day satisfaction dries up.

r_lee 23 hours ago|
and so how exactly are you supposed to provide for your family in this scenario?
jebarker 23 hours ago|||
By having a job. If that job is the same as your intellectual/artistic pursuit then you have to balance the needs of satisfying your employer and what keeps your motivation going over the long term. All I’m saying is that worrying too much about future achievement or “great contributions” are a recipe for burnout and disappointment.
codechicago277 23 hours ago|||
Could work in a patent office or something.
fatih-erikli-cg 10 hours ago||
Nothing. Your personal computer comes with written code. They are written by people who they live in malls. So they deny that they were using your computer and they claim that the written code is math. You pay their fastfood money when you buy a computer. Ask how a square root is calculated if you see mathematican. It is loop that starts from 0 to the answer. Computer does that.
dunham 22 hours ago|
For context - the top answer was written by Bill Thurston, who was awarded a Fields Medal. (Kind of like a nobel prize for mathematics.)
More comments...