STEMfags, what is your actual game plan now?
nothing changed
>>17049879STEM isnt just math and formal systems. It also includes paleontologists in Turkey and social anthropologists that study random Burmese chimp societies or epidemiologists that track some random virus in remote Madagascar island. By the time we have AI robots and models replacing these people, it will be too late for the rest of humanity. I would suggest you informal system STEM fields.
>>17049879if things keep going as they are, death before my next birthday, preferably.
Electrical Engineering remains standing. Sure AI can write verilog or matlab codes now, but anything related to electronics or power systems remain untouched.
>>17049879It is the greatest time for science ever. Only luddites are seething
>>17049893AI can design electronics pretty well. The real know how secret sauce is in manufacturing not design. ASML, Samsung and TSMC are safe but NVIDIA and AVGO eventually will get buttfucked
>>17049896I don't know man AI couldn't help for shit in my electronic courses. Sure frontier models are way better than the crap I used a year earlier, but designing a functional modern circuit requires a lot of know how, heuristic methods, and simulation.
>>17049901I'm not saying it's impossible, it'll just take a long time.
>>17049901>designing a functional modern circuit requires a lot of know how, heuristic methods, and simulation.And proving Navier-Stokes doesn't.
>>17049929Ok nigga go blow 100k on jsut designing an amplifier
>>17049930This guy was all over the epstein files lol
>>17049879Buy as many shares of Infineon, Rambus, ASX, AAOI, Marvell and LITE as I possibly can before I become jobless
>>17049930>smart kids
>>17049930The idea that someone cares about xxAnon420xx's twitter post about how he solved P=NP using ShitGPT 6 is something AI psychotics could conceive of.As soon as serious mathematicians stop cooperating with (((Sam Altman))), (((Dario Amodei))) etc. you and the other tokenslop worshippers are gonna be pissing in a sea of piss, alongside regular schizophrenics who solve dozens of unsolved problems every day.
>>17049896So why OpenAI won't create it own fab?Just use the AI agent!
>>17049879idk but so far the impact of slop on my field has been minimal. I guess no one cares about thin films. You still need a guy who knows the physics of the system to optimize the system and the AI suggestions are kinda silly. Well at least they are kinda good at summarizing stuff but its kinda like jumping from one idea to the next when you only need one step to the right or left.
>>17049879To kill myself
>>17049930This is retarded. A postdoc with the same machine will steamroll a "smart kid". You're going to have to need math understanding to get the most out of these models. By the next few years when all the low hanging results have been neutralized, math will reach higher point where these "smart kids" would themselves need PhD level understanding to get meaningful results out of these models. It's simple really, anything a "smart kid" can contribute, a postdoc would've done it already. Math is shifting to a new co-ordinate system, we still need human mathematicians, their roles will be different.
>>17049882
>>17049882>>17050107Sorry wrong pic.
>>17050104>math will reach higher point where these "smart kids" would themselves need PhD level understanding to get meaningful results out of these modelsYou are really short-sighted. Humans will be needed for nothing. Nothing. Literally nothing. AI can decide what problems to solve by itself. AI can solve those problems by itself. AI will be able to come up with new interesting problems to solve by itself. AI will invent new branches of mathematics by itself. AI will be able to produce dumbed down powerpoint presentations of its own results for dumb humans to ingest. You, humans, will soon not be able to contribute to the research of mathematics in any meaningful way. It doesn't mean that mathematics will be entirely solved, just that like Magnus Carlsen has nothing to teach to the Stockfish Chess engine, current mathematicians have nothing to teach to the AI.
I wish I could mail a bomb or two to Sam Altman
>>17049901>designing a functional modern circuit requires a lot of know howits a discrete exercise... AI will do well to say the least
>>17050208It won't be cheap enough or novel enough
>>17049879Figuring out how to kick out Indians from the internet.
>>17049879Same as it was before. No one I know in STEM cares about the thing you want me to care about. It's STEM-flavored e-celebs and tech scammers putting up a show for American highschool dropouts. The real world continues uninterrupted.
>>17049930Oh, Joscha Bachs, I just last night read about this guy in the wikipedia list of people in the epstein files
>>17050321Look up early life sections and social circle intersections between the following groups:>classic hard sci-fi authors>early technocracy/transhumanism advocates>early founders AI research>current AI CEOs>notorious pedophilesBy the time you're done you will have seen enough cohencidences to before a seasoned schizo.
>>17049879Ropes are cheap.
>>17049929Not if you steal it. Where are all the ither ground breaking proofs? Surely if its that good it could do way more.
>>17049882>nothing changedI'd rather say that every company and university introduced AI training.
>>17049879The only thing most people can do, waiting and seeing what happens. I can't tell how far LLM's can be taken, if they'll ever get enough research taste or judgement to take over the role of researchers completely, but I am steadily less and less impressed by opinions that say LLM's are about to imminently collapse or will never improve because of reason X when I've heard the same thing every year since 2023
>>17050207Today drones are delivering mail, and guess who have all AI for the drones.
>>17049879TMD
AIfags have literally nothing. AI just doesn't work when it interacts with reality. The moment someone tries, everything goes wrong, every single time.
>>17050354I'm steadily less impressed with people saying that the AI is totally good now, given every time they say that they immediately fail to function when it interacts with reality.
>>17049879I know this will sound crazy to read, but if I were an HS graduate I’d NEET/get a low level job for a while. White collar professions are currently too risky to dedicate your time.
>>17050115You can't blame an AI model or robot for faulty research. We need humans at the top for "human quality control" to take moral responsibility of these results especially since math is the bedrock of most science research. Nobody would take Navier Stokes solution seriously if there weren't humans to extract meaning out of it. You send that proof to Sentinel Island and it might as well be chinese witchcraft to them. Otherwise it's just a big game of "trust me bro" with these AI models. As AI keep churning out more and more results, we'd need more mathematicians to keep up with the output, just to hand out "this paper is human verified" stamps like a patent examiner. This is a wonderful time to be a mathematician..not as active researchers but as clerical staff in private AI and govt. research institutions.
>>17049879>what is your actual game plan now?I curbed all my enthusiasm regarding the possibility of meaningful career growth.Also decided to stay away from any form of loans due to job security feeling way too shaky.I still learn new things mostly out of curiosity, but often outside my field these days. Feels like networking/nepotism is more important than ever, and I don't got what it takes to compete in that arena.
apply to corporate jobs, try to get enough money to go to thailand and OD.If that doesn't work I'll probably just kill myself.If that doesn't work out I'll probably get my PhD in cleaning toilets.
>>17049894>this is the greatest time for your replacement ever>only people who care about quality of human life are seething
>>17050604it does seem like moving to thirdie countries will still make for some sort of recognizable day to day life. first world countries are already looking kinda weird on that front
>>17049879Saved up a decent amount of money at my last job so I'll do some travelling, bucket list shit. Then I'll kill myself.
>>17049886>paleontologists in TurkeyIs this some sort of euphemism?
>>17049886>anthropologists >STEMGood one.
you theoretician fuccniggas are cooked, get into nursing, plumbing, day care
>>17050837>changing old peoples' diapers>unclogging toilets>changing little kids' diapersWhy do they all need to involve shit?
>>17050840only sometimes if it's one of those days
>>17049879Teach
>>17049879I will replace AI.
>>17049937>>17050321epstein did nothing wrong. what he did wouldn't even be illegal in most of the world (youngest girl on his island was 14).
