@alz_zyd_ Really? No math? I really think comparative advantage will kick in at some point, even if AI is better at everything (absolute advantage) humans should be relatively better at something.
Successful mathematicians will be those that discover that.
Please resist the urge to mock the mathematicians struggling through this moment. Whatever your field, your expertise, your passion, you too shall one day experience what they are going through. And when you remember your words, you will feel remorse.
I suggest that the right attitude to take toward the OpenAI math release is to imagine we have discovered a notebook full of results from a deceased mathematician. Our job it to make sense of the notebook.
"OpenAI haven’t shared fundamental information about how the problems are selected or the failure rates of their attempts"
This seems unfair. Do human mathematicians have to announce all problems they tried working on and failed to solve?
Yes, but even better. Now every single student and professor has the equivalent of an RA army - a privilege held only by a few senior professors at elite institutions
if you were thinking of applying to phd programs for math/tcs soon you may want to consider econ instead
econ theory has a healthy culture towards valuing new questions, balancing application with elegance, and is well positioned for relevance in "doing math when math is solved"
Concerning math @OpenAI should do the following:
1) Tell us long in advance when they make the next such announcement
2) Release the 4000 problems they tried
3) Try fewer than 4000 problems in their future evaluations and make that list public too
@OpenAI has clearly demonstrated that their models can solve math problems at the highest level. Best for the math community would be if future announcements became much more predictable.
It would be great to know what problems they tried. They also should evaluate their models on fewer problems so that we mathematicians can still discover known questions ourselves. I personally want the models to get better, but it is unnecessary to solve most well-studied serious math problems. Making their future internal benchmarks public and telling us say a month in advance when they make announcements would be strongly beneficial.
Happy to chat with anyone from @OpenAI or other labs about this. Other labs working on math should also do the above.
What are your opinions on this? Do you agree with these recommendations or would you suggest something else?
Solving problems has negative first order effects??
If we didn't want these things solved, why do we call them "problems" in the first place??
Are scholars now anti-researchers, whose role is to limit the pace at which problems are solved???
Insanity!!
Solving problems has negative first order effects??
If we didn't want these things solved, why do we call them "problems" in the first place??
Are scholars now anti-researchers, whose role is to limit the pace at which problems are solved???
Insanity!!
Au contraire.
C'est une nouvelle ère qui s'ouvre pour les mathématiques.
Une ère où la démonstration formelle est largement automatisée et où l'accent sera reporté sur le développement de nouveaux concepts, nouvelles abstractions, nouvelles définitions, et nouvelles conjectures.
L'invention du bateau a réduit l'importance de la nage, mais a permis la découverte de nouvelles terres.
Au contraire.
C'est une nouvelle ère qui s'ouvre pour les mathématiques.
Une ère où la démonstration formelle est largement automatisée et où l'accent sera reporté sur le développement de nouveaux concepts, nouvelles abstractions, nouvelles définitions, et nouvelles conjectures.
L'invention du bateau a réduit l'importance de la nage, mais a permis la découverte de nouvelles terres.
I'm so glad I can think through this storm with some of my favorite tools: opportunity cost, portfolio theory, general equilibrium, welfare theorems, game theory, incentive compatibility, mechanism design, matching theory, incomplete contracts, and asymmetric information
I'm so happy to be an economist in this time of great AI progress and uncertainty.
Tech progress has always come for people's jobs - it is my turn now. Together with engineers, mathematicians, etc.
However, econ gives be two things: perspective and tools.
Perspective: I know this has happened many times - not at this scale, not to these 'type's of people,' but to many people. There is better future ahead.
Tools: Intelligence is now cheaper. This should be analyzed like any other change in relative prices.
It is not from the benevolence of OpenAI, or Anthropic that we expect our mathematical proofs, but from their regard to their own interest. We address ourselves, not to their AI-ness but to their self-love, and never talk to them of our own necessities but of their advantages.
Artists: but I don't want to lose my job. I love it.
Mathematicians: but I don't want to lose my job. I love it.
Economists: LONG HAVE I AWAITED THIS HOUR! DESTROY ME, EFFICIENCY, AND LET UTILIZATION-ADJUSTED TOTAL FACTOR PRODUCTIVITY RISE ON THE WIND OF MY ASHES!