OpenAI's AI Solves 10 Open Math Problems, and Mathematicians Are Coping
Mathematics Without Mathematicians
OpenAI announced that an unreleased model solved ten open problems in mathematics, a milestone that has mathematicians and observers scrambling to rationalize its implications. The author, a software engineer, dismantles common coping strategies—like claiming humans will direct or teach the AIs—arguing that AI will surpass human taste, teaching, and even architectural skills in math. He warns of a future where humans can't understand frontier math or the technology built on it, and compares the situation to chess, noting math's material utility makes AI mathematicians far more disruptive. He concludes with a sobering note on the social context of intellectual work and the choice to obsolete ourselves.
The result is that we will live in a demon-haunted world, full of marvelous devices whose operation we will not understand, based on engineering principles we will not understand, discovered using formalisms we will not understand.
- Twey
> The result is that we will live in a demon-haunted world, full of marvelous devices whose operation we will not understand, based on engineering principles we will not understand, discovered using formalisms we will not understand.
Well, that's just where we started — and then we developed science and mathematics to understand it. Why would it be any different if the demons live in our technological progeny instead of the natural world?
- aureate
Claimed refutation of one of OpenAI's 10 claimed advances (their disproof of Connes' rigidity conjecture): https://philpapers.org/archive/NIEWTC.pdf
I am hopelessly ill qualified to judge.
Edit: Looking around at context, this has major crank warning signs (claims to also prove the conjecture, author has claimed solutions to both the Riemann Hypothesis and the Navier-Stokes problem and is involved in "alternative propulsion" research). I doubt OpenAI would have published without thorough verification.
- RandomLensman
Math might get a lot bigger and then there might be proportionately fewer people but not necessarily in total. Unless humans play no role in economics, business, society, etc., I doubt human intermediation and discovery (or direction of discovery) will go away.
- raver1975
I built a math research engine. If you like reading articles about newly uncovered mathematics, you might like this: https://alethean.org
- c2c2c
I am growing tired of seeing the terms "cope" and "moving the goalposts" in response to discussions around AI. Is there not value in trying to understand what the strengths and weaknesses of the technology are? How can we do that without examining the limitations of the models' current abilities? If we simply dismiss those arguments out of hand as feeble attempts to deny the inevitable rise of superintelligence, then where does that leave us? Shall we all just lay down and start praying to the machine god?