Article says that there are 2 formulations of the NS problem and both are interesting: one is about fluid behaviour with no external forces, other is about fluid behaviour with external forces.
For a counter-example the latter is easier since you can have a tricky external forcefield.
I learned from this charming lo-fi video that it is actually much easier to find singularities in the Navier-Stokes equation for compressible fluids (which is out of scope for the Millenium Prize problem). The first was found in 1998 by Zhouping Xin.
Right - OpenAI proved the forced blow-up case rather than the harder unforced one, with the Millenium Prize problem statement saying it would be awarded for either one.
The forced version is easier since you can custom design the force function to get the result, so getting the blow-up might be regarded just as much a function of your bespoke force function as of the fluid dynamics itself.
My friend, who is a mathematician, sent this in our group chat:
The title is a bit misleading. The variant with a smooth forcing was one of the four valid variants in the Clay formulation. It is interesting to solve it. It is still an interesting and impressive result. The no force version is also interesting and remains unsolved. It isn’t reasonable to just dismiss the proof on the grounds that 26 years later we claim it was never that interesting. This is the first time I’ve seen this attitude.
Not really - if you read beyond the headlines of "mathematicians don't like AI", and your own imagined (incorrect) reasons they may have for that "AI is stupid", then the truth seems a bit more interesting.
If you want to abide by your own words and judge the situation on it's merit, then you need to look at the specifics, meaning the OpenAI proof itself (166 pages), and the analysis of it that is only just beginning. Assuming that the proof is wonderful and provides much insight into Navier-Stokes is just as dumb a take as assuming that it doesn't. Judge it on its merit.
The Scientific American article is sadly paywalled, but at least part of the discussion is based on the paper below, whose work the OpenAI proof appears to build upon.
When discussing the proof itself, below, with Sonnet, and asking it to explain the distinction between a function being smooth and analytic, one aspect that appears interesting is that the OpenAI forcing function is apparently constructed out of "bump functions", meaning that it is not a uniform force acting upon the flow but rather a highly engineered pattern of pokes, localized in time and space, which as another commenter in this thread notes sounds similar to Maxwell's Demon - another theoretical force, that neither tells us anything about Brownian motion nor the 2nd "law" of thermodynamics.
So, we'll have to wait for mathematicians to continue to analyze the proof, and determine to what extent is does deliver on providing insights into Navier Stokes, and any potential improvement to it, that was the goal of setting it as a Millenium Prize in the first place.
So what they are saying is that humans, in this case Charles Fefferman (a math prodigy, going by his history), failed to specify the problem correctly?
Yes. And more humans—in this case OpenAI researchers—similarly failed in choosing how to direct the AI tools.
And yet another set of humans—Open AI marketers—made an error in how they sold the result of the preceding errors.
But its not news that computers are mere tools and that any error blamed on a computer involves at least two human errors, one of which is blaming the computer instead of the human(s) responsible.
Its perhaps a bit less obvious that every thing for which credit is given to a computer involves at least one human error—that of crediting the computer—and certainly can be more amusing when it involves a bunch of human errors.
Fefferman was right to include option C as someone could have come up with less contrived counterexample accompanied by some interesting theorems that actually shed light on the general case
Doesn't the way OAI's proof is formulated essentially rule that out, in that as long as an alternative solution concerns option C, it'd have to live in this same solution space they carved out?
No, OpenAI did not solve the "wrong" Navier-Stokes problem. OpenAI did not solve the hardest version of the problem (unforced blow-up), but did give a solution to the Clay Millennium Prize Problem as written and understood, choosing the explicitly allowed forced option.
SciAm writes "in a sense, the LLM found and exploited a loophole in the framing of the question". This is pure sensationalism. Choosing option (C) (out of an explicit list of four options) is neither a "loophole" nor something "found by the LLM"; everyone involved knew this was the option they were pursuing.
With the grumbling out the way, there is some actual scientific content to the article: there's a strong argument that OpenAI's method will not extend to the unforced case, leaving our understanding of NS incomplete. This negative result is itself new and interesting (and predicated entirely on the solution found by OpenAI)!
Totally agree. I wouldn't quite call it clickbait, but the article does this thing I find annoying where it puts the "sensationalist" framing at the beginning (the "loophole" quote you put), but then closer to the end fully admits that it wasn't really a loophole in any case:
> It did, however, unambiguously solve the problem according to the Clay Institute’s original formulation. The official problem statement, penned in 2000 by mathematician Charles Fefferman, offers an option called “C,” in which solutions are allowed to use an external force like OpenAI’s.
As I understand it the "loophole", if you want to call it that, is that OpenAI's custom-designed forcing function was smooth, as the rules said it had to be, but was non-analytic, consisting of some construction of "compactly supported bump functions", meaning a mass of tiny little pushes at precise points of space and time to push a vortex into blowing up the math.
> SciAm writes "in a sense, the LLM found and exploited a loophole in the framing of the question".
God, it’s embarrassing to read stuff like this. They’re making it seem as if everyone involved was either stupid or dishonest just so they can pretend they have a scoop here.
There's a big market in journalism to take down the popular thing in the news. "Everyone is wrong" and here is my article where I wildly exaggerates some minor details to justify the headline which made you click on the article.
Not really. All of the Navier-Stokes options in the Clay Institute formulation of the prize are real problems of significant interest. The history of this particular problem is of finding specific conditions under which we can get something to work that turn out not to generalize in ways that people don’t expect eg iirc (It’s been a while since I read about it) there was a solution found early-ish in the 20th century for the 2-D case that turned out not to generalize to N-d, N>2 case, there are special conditions under which the turbulent terms cancel out and you can get smooth flow, vortices etc.
It seems to me that formulating problems at the boundary of human knowledge precisely is always going to be challenging and situations are bound to occur where you look back with the benefit of hindsight and wish that you had posed the question slightly differently based on some knowledge you didn’t have at the time.
It's just moving the goalposts, this happens every time an AI solves a problem, doesn't matter if the goalposts were there for 26 years.
What's interesting is that there are a set of people who are "in charge" and can as they wish arbitrarily set the goalposts to the thing that they happen to be best at. While this might be satisfying for an Humanity vs AI narrative, it's concerning for an us vs them one. Are these people really special? or do they just change the rules of the game so that outsiders (human or AI) can't win.
“Yes they solved one of the 7 most famous unsolved problems in math today but they only did the easiest version!”
At the current rate (if they keep burning tokens on it, which maybe they won’t given the backlash) RH will be proven within a year and there will be some other thing that means it’s not actually that impressive…
This situation illustrates exactly the limitation of AI and why we still need humans in the loop.
It reminds me of a junior coding bootcamp lecture I once gave many years ago before AI coding. One of the first slides said "Computers will do exactly what you say, not what you mean."
When the news spread about the solution of this problem by AI we started wondering what will happen when AI will start generating proofs we can’t comprehend.
Today, we are discussing if AI cheated by picking the easy problem to solve which means that we at least still comprehend what’s going on.
I wish mathematics and the rest of the human intellect wouldn’t turn into content marketing that is generated primarily to trigger strong human emotions.
I feel that this is going to hurt both AI and the disciplines that can benefit the most from it
The open question is whether the Clay institute will award OpenAI or others the Millenium prize for this solution to the Navier Stokes problem. I speculate that they won't. OpenAI's solution is undergoing peer-review and the counterargument presented in this article changed my mind. By default, if the Clay institute hasn't officially recognized the solution as true or likely true, I can't make an assumption that it is. I hope to go through the proof and perhaps AI can help better understand it and any potential weaknesses.
There might be more value in having LLMs systematically hunt for mistakes in existing and widely assumed correct math papers, or hunt for counter-examples to things thought proven. There's likely to be a couple mistakes hiding in the less well scrutinized edges of mathematics.
Maybe something interesting will fall over because of that, who knows?
39 comments
[ 0.23 ms ] story [ 24.3 ms ] threadFor a counter-example the latter is easier since you can have a tricky external forcefield.
https://www.youtube.com/watch?v=4wEn9B7pDV4
The forced version is easier since you can custom design the force function to get the result, so getting the blow-up might be regarded just as much a function of your bespoke force function as of the fluid dynamics itself.
The title is a bit misleading. The variant with a smooth forcing was one of the four valid variants in the Clay formulation. It is interesting to solve it. It is still an interesting and impressive result. The no force version is also interesting and remains unsolved. It isn’t reasonable to just dismiss the proof on the grounds that 26 years later we claim it was never that interesting. This is the first time I’ve seen this attitude.
Almost never do people judge situations entirely on merit.
If you want to abide by your own words and judge the situation on it's merit, then you need to look at the specifics, meaning the OpenAI proof itself (166 pages), and the analysis of it that is only just beginning. Assuming that the proof is wonderful and provides much insight into Navier-Stokes is just as dumb a take as assuming that it doesn't. Judge it on its merit.
The Scientific American article is sadly paywalled, but at least part of the discussion is based on the paper below, whose work the OpenAI proof appears to build upon.
https://arxiv.org/pdf/2609.20803
When discussing the proof itself, below, with Sonnet, and asking it to explain the distinction between a function being smooth and analytic, one aspect that appears interesting is that the OpenAI forcing function is apparently constructed out of "bump functions", meaning that it is not a uniform force acting upon the flow but rather a highly engineered pattern of pokes, localized in time and space, which as another commenter in this thread notes sounds similar to Maxwell's Demon - another theoretical force, that neither tells us anything about Brownian motion nor the 2nd "law" of thermodynamics.
So, we'll have to wait for mathematicians to continue to analyze the proof, and determine to what extent is does deliver on providing insights into Navier Stokes, and any potential improvement to it, that was the goal of setting it as a Millenium Prize in the first place.
https://cdn.openai.com/pdf/32d9f210-8b73-45e0-91bc-82a30aef8...
And yet another set of humans—Open AI marketers—made an error in how they sold the result of the preceding errors.
But its not news that computers are mere tools and that any error blamed on a computer involves at least two human errors, one of which is blaming the computer instead of the human(s) responsible.
Its perhaps a bit less obvious that every thing for which credit is given to a computer involves at least one human error—that of crediting the computer—and certainly can be more amusing when it involves a bunch of human errors.
Next year's headline: But did Skynet kill all humans?
SciAm writes "in a sense, the LLM found and exploited a loophole in the framing of the question". This is pure sensationalism. Choosing option (C) (out of an explicit list of four options) is neither a "loophole" nor something "found by the LLM"; everyone involved knew this was the option they were pursuing.
With the grumbling out the way, there is some actual scientific content to the article: there's a strong argument that OpenAI's method will not extend to the unforced case, leaving our understanding of NS incomplete. This negative result is itself new and interesting (and predicated entirely on the solution found by OpenAI)!
> It did, however, unambiguously solve the problem according to the Clay Institute’s original formulation. The official problem statement, penned in 2000 by mathematician Charles Fefferman, offers an option called “C,” in which solutions are allowed to use an external force like OpenAI’s.
God, it’s embarrassing to read stuff like this. They’re making it seem as if everyone involved was either stupid or dishonest just so they can pretend they have a scoop here.
It seems to me that formulating problems at the boundary of human knowledge precisely is always going to be challenging and situations are bound to occur where you look back with the benefit of hindsight and wish that you had posed the question slightly differently based on some knowledge you didn’t have at the time.
What's interesting is that there are a set of people who are "in charge" and can as they wish arbitrarily set the goalposts to the thing that they happen to be best at. While this might be satisfying for an Humanity vs AI narrative, it's concerning for an us vs them one. Are these people really special? or do they just change the rules of the game so that outsiders (human or AI) can't win.
At the current rate (if they keep burning tokens on it, which maybe they won’t given the backlash) RH will be proven within a year and there will be some other thing that means it’s not actually that impressive…
It reminds me of a junior coding bootcamp lecture I once gave many years ago before AI coding. One of the first slides said "Computers will do exactly what you say, not what you mean."
Today, we are discussing if AI cheated by picking the easy problem to solve which means that we at least still comprehend what’s going on.
I wish mathematics and the rest of the human intellect wouldn’t turn into content marketing that is generated primarily to trigger strong human emotions.
I feel that this is going to hurt both AI and the disciplines that can benefit the most from it
OAI should just get on with it and more importantly - produce more stuff that positively benefits the vast majority of the population.
Maybe something interesting will fall over because of that, who knows?
We should either see the effect or get to understanding of how to correct the model.
Note that the force is finite and doesn't "know" the vortex config. The singularity should be achieved in a finite time.