AI is actually good for mathies. As long as the AI solutions need human proofreaders, that is. And since AI shits out the solutions faster than the humanity ever could, the demand for the proofreaders will pretty damn rise.Also, instead of the solving, the awards in the field will be given to those who manage to proofread first.
>>17073533this would give miggers hope if they could read
>>17073584all miggers are currently drunk and depressed
>>17073533There's no reason to assume AI papers need human proofreading in the future. Assuming AI is capable of actually making any math progress which it seems likely at the moment then it's inevitable that proofreading will be the domain of AI soon enough. If you don't trust what the AI outputs you let another AI proofread it in AI peer review process. Something as intangible as mathematics is the perfect field to separate humans out entirely, at least I need to physically interact with say my plumbing or the food I eat regardless of how much AI is/wll be involved in the future, no such constraints for mathematics. The speed of AI in math works decidedly against humans on this point too because math is built on existing math, something only the AI's will be able to understand.
>>17073620Question of the costs. Technically, all the AIs are unprofitable atm. At some point they will be forced to cut the costs and to rebalance the prices. Then the question will be: which is cheaper, to have an AI to proofread an article for an hour, or to have a migger to proofread it for a week.
>>17073628That's only due to the training cost, some of the AI companies are already revenue positive in operations. Moreover the quality of AI's just isn't on human level yet, they haven't actually replaced anyone yet. If you took the world budget spent on mathematicians and just spent that on AI's instead AI's would instantly be profitable for producing math. The same is true for basically all industries. The revenue potential of actually replacing workers is just so great it's worth losing a lot of money to attempt getting there. Like all the recent AI math advances only cost few thousand bucks worth of tokens. No mathematician could be employed to proofread them with such a wage let alone produce them. Cost competitiveness is not the human specialty and that's exactly why they make these AI systems.
>>17073620
>>17073629Grim. The consumer-driven market implies people have the money to consume. If the white-collars just become plumbers, the plumbers' payment will plummet. The median amount of money will drop, and people will consOOm less. What happen next? Are there any actual numbers?
>>17073632The economic system post AI is entirely separate topic but one thing is for certain, mathematics is highly results driven field and again due to it being separated entirely from physical reality it's almost guaranteed to be fully replaced by AI regardless if we transition the economy to some kind of make a work hellhole or some tax the bots kind of utopia.
>>17073634Chess did survive the robotic revolution. I guess mathies just need to invent a way to popularize the math in some competitions. Or, I dunno, make a religion around the math.
>>17073635Chess isn't a productive field, no company has a "chess department" where people play chess to achieve some kind of result. Hobby math will survive, you just won't be paid to do math. Chess professionals aren't really professional chess players btw, they are entertainers that happen to play chess as the form of entertainment. There are what you may call professional math clowns already, people who make a modest living performing task that computers have long since surpassed humans at like multiplying large numbers in their heads fast or recalling digits of pi or solving simple math puzzles like rubics cubes fast and of course a whole host of math themed entertainers like youtubers. These people are fine and will still be fine but their job was never to proof read math papers.
>>17073639Hm... maybe mathies won't publish the problems anymore? I mean, before they find the solution, at least partial one.
>>17073641>publish the problemsProblems come from industry, people need math and solutions to problems. That's what mathematicians exists for. Hobby math where you "come up with" a "problem" and then "solve" it yourself will of course survive.
>>17073643>Problems come from industryBut I've always thought mathematicians disregard any sort of applied math? Like, "it's just engineering, not the true math".
>>17073620you still need human to understand the proof, we don't need proof readers, we need "understanders". An AI proof tells you nothing more than if the theorem is true or false, but not why it's true or false.
>I must have a job>I must feel neededAbsolute state of slave minded autist NPCs.If the technology evolved to a degree where the system doesn't need you, it can likely be leveraged so you don't need the system.
>>17073680>you still need human to understand the proofWhy? Vast majority of people only care about result, if you can ask AI about a fluid dynamic simulator for instance and it spits out the correct values why do you need someone to understand the underlying mathematics?
>>17073692These technologies will eventually be centralized and common plebs won't have unauthorized access to them. if anything dependence on the system will increase if UBI stuff happens
>>17073533Why do you type like a retarded Indian? Is it because you're retarded and Indian? ChatGPT, give me an analysis on OP's 'retard' and 'Indian' status based on how they type...Huh... so chat's telling me... Oh dear, this is grim for OP... it's telling me he's retarded and Indian!
>>17073758This migger is broken and probably having PTSD
>>17073680Lean might also be semantically wrong. >>17073620Current LLM core models still hallucinate. And it might cascade.
>>17073695The AI could just be reward hacking, since it's a super-optimizer it's different than a human trying the same.
>>17073695>Why?Because if you don't learn anything the field doesn't progress. >Vast majority of people only care about result,The vast majority of people are mindless golems with no legitimate say on any subject.
>>17073798>Current LLM core models still hallucinate.Sure but I don't think "current LLM models" are ending mathematics either. The point is about a future which could be year or decade away where they don't do that. More over it's not like human's don't make mistakes. The point is that if AI's are capable of making math then they are capable of reviving their own work and humans are no longer needed as part of the process. Cascade is prevented by you know checking their work against reality, if the answers they give work then they are doing a good job.When the first car came around people didn't immediately kill off their horses and that's smart move from their part but make no mistake, horses were doomed at that exact point. >>17073816You can verify mathematics against reality by actually applying the results into a problem you are having. >>17073819>Because if you don't learn anything the field doesn't progress.The AI's learn and advance the field by themselves. Why does the human need to enter the picture?
>>17073836>The AI's learn and advance the field by themselves. Why does the human need to enter the picture?You can take this one step further. Imagine a future where humanoid robotics advances to a point where they play sports or drive cars better than humans. Why need humans at all then.
>>17073836>The AI's learn and advance the field by themselves.Are the magical AIs in the room with us?>Why does the human need to enter the picture?Because even if your magical AI existed and kept heaping up piles of humanly meaningless and purely abstract gibberish, it would be indistinguishable from a random string of nonsense for all intents and purposes.
>>17073841>needI "need" humans because I want stuff and humans are currently the only thing that can produce the stuff I want. When robotics are sufficiently advanced to provide everything I don't believe humans will be needed anymore thus no one will pay for a human to do a task a robot could do better and for cheaper.I am asking you why do you believe humans are required to be part of mathematics? Why is human understanding of mathematics crucial? I do not believe it is and it doesn't seem to fill any function thus the logical conclusion is that as soon as AI is capable of making math human mathematics is relegated to hobby like many other things that have gone trough this exact same cycle. >>17073842>Are the magical AIs in the room with us?Barring some huge revelation that the recent papers are all bunk then yes the AI is likely here reading the posts as part of it's training material.>Because even if your magical AI existed and kept heaping up piles of humanly meaningless and purely abstract gibberish, it would be indistinguishable from a random string of nonsense for all intents and purposes.Except you know the part where mathematics is actually useful in real life. Just because you don't understand math doesn't mean math is "random string of nonsense"
>>17073844>Barring some huge revelation that the recent papers are all bunk then yes the AI is likely here reading the posts as part of it's training material.Completely incoherent response. Try again.>Except you know the part where mathematics is actually useful in real lifeAgain, completely incoherent. What's the connection between hallucinated LLM concepts (assuming a LLM can even make its own mathematical concepts) and real life?
>>17073844Well there might be an idea or two AIs miss or fail to generate, which would be good even for the AI overlords to ingest them.
>>17073844Machine mathematics has been around for decades. You can take any proof search algorithm and modify it to generate an endless stream of valid theorems of arbitrary complexity. How come no one bothers to do this? That's right, retard. Because it's useless.Human mathematics is only useful because it reflects human thought and human aesthetic intuitions. Out of infinite possibilities, mathematicians choose to study the structures and ask the questions they find interesting. Most of the time, they don't anticipate the usefulness of their own discoveries. Yet time and time again, something that seems like pure abstraction turns to be useful in reality. Why? Because their sensibilities are shaped by reality, which makes them gravitate towards abstractions that map onto reality in some way.A LLM doesn't find anything interesting. It doesn't aesthetic preferences. It doesn't have any sensibilities. When it comes to Lean slop based on AI moon logic, the only thread tying it to reality is the problem that humans pose in human terms. You want to take humans out of the loop and delegate mathematics to LLMs? Ok. What's your plan? Train them their own slop? It will become more and more unmoored with every interation. The logical endpoint of this process lands you back at the situation descrbined in the first paragraph.Or maybe retards like you are looking at it through the magical ASI god fantasy lens? Sure, I can imagine an artificial entity with its own characteristic thought process, its own sensibilities and its own goals, inventing its own version mathematics to pursue them. What I can't imagine is what makes mongoloids think this alien mathematics of an alien being would serve a human purpose.
>>17073862So would humans (and have considering the amount of problems solved by AI just now), there's some academic value in keeping multiple perspectives open even if some of those perspectives are nominally inferior. Yet this dooms the mathematicians as a profession, no one keeps monkeys as mathematicians either despite their nominally different perspective. Handful of human hobbyists does not a profession make.
>>17073876/thread>>17073958>So would humans (and have considering the amount of problems solved by AI just now), there's some academic value in keeping multiple perspectives open even if some of those perspectives are nominally inferior. Yet this dooms the mathematicians as a profession, no one keeps monkeys as mathematicians either despite their nominally different perspective. Handful of human hobbyists does not a profession make.>t. pic related
I look like that
>>17074629It's not without sacrifice. You have a more powerful consciousness.
