It seems the AI skeptics have just decidedly to collectively ignore this and pray it doesn't happen again
>>108815628Surely this guy has proof of this correct?
>>108815628>mathematics departments, who owe a duty of care to their students, should be urgently preparing for it
>>108815655https://gowers.wordpress.com/2026/05/08/a-recent-experience-with-chatgpt-5-5-pro/He's a Field's medalist btw>>108815669I think you mean nursing.
>>108815628Is he saying that AI actually made up something noteworthy or that it simply generated a reasonable chapter that could be put into a PhD thesis? Because I don't care about the latter point.
>>108815681What is he invested in? Does he or does he not have a financial incentive? Is there any proof the AI actually did what he said it did?
>>108815693It proved a result
>>108815702proved what result?
>>108815713https://drive.google.com/file/d/11r-ggU__GMmHIrgEHQVULUIR1VxKSwmi/view
>>108815628if you are not using AI to speedrun through a degree you are retarded>inb4 uncs kvetching because they had to endure the humiliation ritual of pointless exams
>>108815669I literally CANNOT believe that anyone with a working brain can't learn everything needed to become a plumber in a week and the rest is simply learning on the job and be willing to be elbows deep in other people's shit. Change my mind.
>>108816867Most plumbers are 80-90 IQ so learning how to turn a wrench for them is hard. The real skill is from fleecing clueless customers.
>>108815628Phd theses are slop buckets that nobody reads, not even the "referees". It's not surprising that the slopbot can make slop.
>>108815628>it's another episode of technically illiterate AI users thinking that if they didn't have information that gives away the solution to a problem, that information definitely isn't among the 500 TB of data token shitters were trained on
>>108815628>AI skepticswtf is even your point tho? That AI is sentient AGI?
>>108815628math is literally just symbol manipulation so naturally LLMs are good at it just like language and code
>>108815628Snailcat luddites BTFO once againThey will be replaced and destroyed. We vibeGODS won
>>108816867You can learn to conform to building code in a week, but the reality of being any kind of tradie is dealing with bizarre anachronistic stuff that isn't in the manual and all the nuances of dealing with improperly installed or falling apart equipment. Which they pretty much have to learn through experience because guys with experience are either not ambitious writers or not interested in undermining their successor's racket.Doubly so with plumbers because they're dealing with raw sewage under pressure and the consequences of fucking up are dire.
>>108815628This topic always provides lots of evidence that this bird is extremely retarded
>>108817802Why'd you have to paint on the Nazi paraphernalia? Couldn't you bobot have done it for you?
>>108818472My brain is plugged into a computer running Claude mythos, therefore it was vibeDRAWN. You wouldn’t get it ludder. We don’t need you anymore
>>108815695Isn't everyone employed in the US invested in AI one way or another?
>>108818487There's a difference between diversified in and invested. It makes no difference to the average working person whether one class of assets goes up or down. That's the fund manager's problem.
>>108818487Gowers is British and has a chair at Cambridge…
>>108815681>>108818462Basically, Nathanson and Rajagopal are in an esoteric subdiscipline sniffing their own farts, since there is nobody else in their esoteric subdiscipline.The parrot cited Rajagopal as its source, cleanly entering the circlejerk.At best, the parrot helped them organize their thoughts. At worst, the parrot is saying nonsense but Rajagopal can't help but sniff his own farts.The student still has no clue what is going on, and because his mentors are fart sniffers, assumes they are right because they taught him to sniff farts also.
>>108815628it's humorous how you say "lmao look at the luddites", then turn around and go "-but AGI going rogue and exterminating us is just scifi".an idiot laughs at himself in the past.
>>108818331Yes, but if large parts of white collar work gets automated before manual work, it's really short sighted to think the trades are safe if the hump before being able to get started is so small. If you have are prudent and methodical, you can probably do most things without causing huge disasters even while you're learning. And as terrible as most YouTube tutorials are, I'm sure there are some perfectly ok ones that can teach people a lot of what used to be only possible to learn by experience.If white collar workers are fucked, tradies are too and unions won't save them.
>>108816867The thing about learning through working is you learn most from your mistakes. Water is a bitch though and a mistake might lead to millions worth of damage depending on the property and the scale of the fuckup. You want your plumper to know his shit but I understand that for an office monkey with zero practical knowledge this is hard to grasp.
>>108819390I think you're underestimating what the average non-retarded person can do if they have to.
>>108816933>If you only knew how bad things really are
>>108815628I don't know enough about PhD level math to personally verify that the AI's contribution was actually meaningful and absolutely everyone who might confirm or deny it has every reason to lie to me.I'm going to continue suspending my judgement until I can find an opinion I can trust about it or until I get a PhD in math.
>>108815628Is the real question: Do we inherently need PhD students? Like for real, not talking about peoples feelings.
>>108815628This sort of advanced maths research is completely useless in itself. Literally the whole point is to train very smart people who can go in other fields and do something useful. If you have a LLM do it, then what is the point?
>>108819687Yes, because PhD students are slave labor. When you consider things like drug development, 50-75% of the work is done in academic labs which need boots on the ground to do the pipetting (that's PhD students). A more important question is: do we need mathematicians (outside statisticians which mathfags consider to not be mathematicians anyway) and grad mathematicians in particular. The answer to that is a very obvious "no". Same with modern physicists: they've been in over their heads for the past few decades with 0 useful results (only fraud). They still can't crack fucking optics despite the trillions put into it either. Now it's all strings this and dark matter that, all of it is just mental masturbation and fraud unironically.
>>108819797Mathematicians are totally incapable of doing anything outside of math, though. Anytime something useful is found in math, it gets cast off as 'not real math'. Computer science, for example. Hence, if you think 'useful' = 'cs' because the og computationalists had math degrees, you fail to understand the context: cs degrees used to be part of math departments until what, 40 years ago? But they really weren't doing math. The current cs curricula are far more mathematically rigorous than they were in those days.
>>108815628I dont think there is a mechanism in the college system that prevents someone's PHD thesis from covering the same ideas as PHP thesis from other students.AI just mashes together one for him not because of any new ideas or concepts (llms can't actually think or apply knowledge), but because there are enough PHD papers in the dataset covering the things he wanted to see that it was able to produce (plagiarize) one just like them.
>>108819687Academia exists as a clerical system to let people live off by accumulation of study. Degrees having value as a seal of approval for other sectors is a byproduct.
>>108819857>Yes, because PhD students are slave labor.Came to write this.>>108819687Yes. Check out >>>/sci/scg and read the FAQ. Doing a PhD is slavery, doing a postdoc is life in poverty. It was recently mentioned that a postdoc in the UK earns as much as a burger flipper at McDonalds.
>>108820646>It was recently mentioned that a postdoc in the UK earns as much as a burger flipper at McDonalds.This isn't the whole picture either, as postdocs and PhDs are expected to work very very long hours. In many schemes, they earn a grant rather than a salary, which means they are denied social safety nets. For postdocs, the contract is usually one year renewable, so they basically immediately have to start looking for the next postdoc opportunity as soon as they start, on top of also doing long hours at work. All this on a burger flipper 40 hour equivalent wage.
