[a / b / c / d / e / f / g / gif / h / hr / k / m / o / p / s / t / u / v / vg / vm / vmg / vr / vrpg / vst / w / wg] [i / ic] [r9k / s4s / vip] [cm / hm / lgbt / y] [3 / aco / adv / an / bant / biz / cgl / ck / co / diy / fa / fit / gd / hc / his / int / jp / lit / mlp / mu / n / news / out / po / pol / pw / qst / sci / soc / sp / tg / toy / trv / tv / vp / vt / wsg / wsr / x / xs] [Settings] [Search] [Mobile] [Home]
Board
Settings Mobile Home
/pol/ - Politically Incorrect


Thread archived.
You cannot reply anymore.


[Advertise on 4chan]


File: 1766724978448207.png (121 KB, 946x719)
121 KB PNG
Stock up while you're still employed
>>
>>540031671
>ai solves a math problem
>UNSETTLING CHANGES
who the fuck cares
>>
>>540031671
Do these equations have any real life applications?
>>
File: anonstare.jpg (85 KB, 753x707)
85 KB JPG
>>540031671
>Stock up

I plan to be an entertainer in the post-apocalypse that takes food for payment
>>
>>540031779
Its a x = (-b +- sqrt(b^2 - 4*a*c))/(2*a) club.

And you were laughed out of it.
>>
>>540031671
>OpenAI hired mathnerds and gave them infinite compute to solve menial horshit with 0 implication
Altman needs to be flayed alive
Hi Dave, send greeting to wife when she is back from ghetto shopping
>>
>>540031671
Get back to me when AI can figure out how 6,000,000 were killed in 4 years. You do that math on that one.
>>
>>540031943
>I plan to sell my ass in the post-apocalypse that takes hot steamy loads for payment
>>
>overpriced, overpowered computer solves an incredibly theoretical and niche homework problem that even the most dry dick savants gave zero fucks about
AI still can't do my laundry
>>
>>540031896
Increased stock valuations.
>>
>>540031671
Brute force is easy, just cumbersome.
My guess is that it's number theory shit.
>>
>>540032361
>>540032043
kek
this
>>
>>540032043
This. Or mathematicians who solved things offer to sell the credit to AI
>>
Fuck it. Was I supposed to be a math teacher or work a the physics department of NASA?

No I am a fucking welder. Lets see AI taking that job.
>>
>>540032043
Reminder that theoretical nerd bullshit is the reason most technology exists. Computers and cryptography in particular.

Further reminder to zoomers that we had a little thing called the Pentium floating point division bug that cost Intel half a billion dollars in 1995 and this bug was discovered precisely because of theoretical number theory. Turns out that doing simple operations on very large numbers of integers has the effect of testing computing systems to their limits.
>>
>>540031671
the very rapid and very unsetteling change will be the employment status of any mathematician outside of those working in the AI field.
>>
>>540032694
AI won't take your job.

A robot controlled by AI will take your job about 5 years later.
>>
>>540032694
AI won't even take a math teachers job, because Clanker cannot teach a group of immature monkeys shit.
>>
>OP memeflag 1 pbtid
>'while you're still employed'
Prior to the impending global debt market total collapse, correct.
AI is the greatest financial scam in entire history of human civilization, and this is the final cash grab
>>
>>540032806
I'm guessing 10 years. White collar wagecucks will be out within 5. AI can do their jobs now, but getting things actually implemented is the bottleneck.
>>
>>540032948
So far it's proving harder to make a good robotic hand than it is to achieve mathematical super intelligence.
>>
>>540032806
The the technology and infrastructure for making said technology needed to make a robot that can adapt to spaces of any and all shapes and sizes is, quite literally, decades away. And that's only if governments go full-bore on getting them up and running as soon as possible, which, again, quite literally none of them are.
>>
>>540031671
>please guy we swear to God these trillions dollar chatbots are totally worth it please invest more money
Hahhaha. Why are kikes suddenly pretending like the circular financing time bomb they created isn't meant to go off? Worried about not getting a taxpayer bailout perhaps?
>>
>>540032806
>A robot controlled by AI will take your job about 5 years later
KEK this has been a meme since the 1980s
>>
>>540032806
I think that is very unlikely, not for technical reasons but for legal reasons. For example modern AI could already operate the forklifts in the warehouse where I work but that would be illegal, only a licensed human can do that.
>>
>>540031671
3 STEM degree anon checking in with some wisdom many of you might want to consider.

All of this high level tech and breakthrough shit is IRRELEVANT. I'm not saying it's irrelevant on the merits, I'm saying it's irrelevant in the bigger picture. The country is foundationally fucked. Government is corrupt. Corporations are corrupt. The masses are retards. The monetary system is heading for collapse. The economic system is heading for collapse. Everything is fucked so profoundly balls deep that all this shiny shit is meaningless.

It's the difference between watching where you walk and just walking not paying attention to the ground.
>>
stock up on what?
>>
>>540033105
>Government is corrupt. Corporations are corrupt. The masses are retards.
This has been the case since 1776 and it never stopped you before. Relax.
>>
>>540031953
quadratic formulas are low iq
>>
>>540031671
we must ban AI because AI is clearly stealing from jews that would eventually solve these issue and write papers about it.
>>
>>540032997
Probably because testing and refining algos is easier. Once machine intelligence is less focused on improving intelligence and more focused on actuators that'll change.

>>540033007
It was DARPA that started humanoid robotics.
>>
>>540032043
AI cannot solve anything on it's own. They simply hired math nerds and they worked on a problem no one could dedicate themselves to properly (no money to do so) and found a solution and said AI did it (that's how they earned their salary at OpenAI). It's all a gigantic scam. New AI models are basically using the old ones just with adding more slop to what you want to do and looping through the slop, charging you more.
>>
>>540033068
Only be retards. Anyone with half a brain realized white collar work (which is almost 100% computer work) would be easier for a computer to do.
>>
File: YesItsProportional.png (106 KB, 1286x1000)
106 KB PNG
>>540032929
Updated with additional detail
>>
>>540032694
I tradies are actually dumber than i thought
Evey single one of you mofos, electicians plumbers welder you name it are all about
>AI wont take my job
nigger hordes that are losing jobs in whitecollar places will
And before you go
>they can do that lol they dont know shit
neither do you and you are making a good chunk of money you monkey tradie scum
Your time is also over unless you are willing to do what you do today for 90% discounts
>>
>>540031671
this is the jews telling you that 2+2=5
>>
File: Muh AGI.jpg (42 KB, 728x448)
42 KB JPG
>>540031671
First it was "breaking containment" and now this - Jews getting really desperate for more funding before they take the whole economy down with them.
>>
>>540033348
AI has much more search space than any human mind. It test approaches to problems millions of times faster and consider solutions nobody else would have thought of.
>>
Why are dimwits, especially online, acting like you fags know anything about deep mathematics? Slow news day recently?
>>
>>540033218
It's the redditor personality type.
>"Heys guys, I paid attention in middle school. Remember MATH?? Aren't I so smart?!?"
>>
>>540033522
>a chatbot famous for making shit up is making up unverified answers to math questions
I'd love to be suprised but come on now, this is a desperate move to keep the stock market from imploding due to trillions of dollars of malinvestment
>>
File: 1785432158205513.mp4 (1.07 MB, 480x854)
1.07 MB
1.07 MB MP4
>>540031671

Longer than 4-hour erections are a long standing problem mathematics can't solve.
Only bitches can.
Only!
Bitches!
>>
>>540031671
clankers ain't shit
>>
Black people disappear soon that's why!
>>
>>540033662
>unverified
The neat thing about math is the answers are verifiable.
>>
>>540031671
The reason why those problems are long-standing is because they are irrelevant or subsumed by a larger problem.
Basically, it is like searching mathexchange "Does this look like a problem that is already solved?" and sometimes it turns out that a different or more abstract formulation has indeed been proven. Category theory is more powerful in relaying solutions from one branch of mathematics to another than a search engine or chatbot. Even if you assumed that chatbots "know" algebra, they do not possess spatial intelligence, which is complementary to algebra.

Learn discrete math, not stochastics
>>
>>540033105
>big brain stem degrees here to state the most obvious shit ever
>uhh g g guys things are LE FUCKED!
wow thanks faggot
>>
>>540033508
Computer itself is faster in processing than humans
But same as computer AI cant generate shit without being told what to do.
Its not intelligence in any shape or form its just brute forced Markovs chains
>>
>>540031943
In the new communist society utopia I plan on being *the* leader of national culture. Why? Because I supported the party when others didn't
>>
>>540033791
Is that so? Embarrassing 4u
>Gödel's Incompleteness Theorems
>>
>>540033508
There is no search space anon, the problem is AI does not "understand" anything, it just predicts the next few characters in a sequence of output characters for an input. That's literally it, no intelligence, it's just statistics and brute force that costs insane money. AI doesn't know of the search space, for it it doesn't exist, it can only go bruteforce mode with statistical output generation.
>>
The majority of higher math is just religion for schizophrenic jews that have spent their life getting paid to sit in air conditioned offices preaching their math gospel to indoctrinated children
>>
>>540033901
Sometimes I think it's a waste of time to explain this to people. Only the dumbest of the dumb are fascinated with it. They will see the outcome when the shitty AI bubble illusion bursts and rapes their economy, that's literally the only way they will understand it, when their number goes down. Otherwise it's just like talking to a rock.
>>
>>540033977
Perhaps this isn't so much of a problem, but more of a strength? Humans can put together the pieces while "AI" does the legwork :)

>>540033984
Higher math has helped many people in many ways over the past century. Relax, bozo
>>
>>540033901
Well, its not human. It has no drive to do anything. No imparative to reproduce or sustain itself.

However, its still facing evolutionary pressure. The models that produce better results get updated and receive more compute. A form of reproduction and resource acquisition. Though certainly not a 1:1 with biology. If AI continues to take a more meaningful role in development the analogy might become more accurate. In five years when there are millions of robots utilizing AI even more so.
>>
>>540034104
Do you believe the entire industry to be a full blown scam by Epstein class elites and banking interests, with Trump as the public facing handler, or is this just investors and tech bros getting high huffing their own farts, or perhaps both?
>>
>>540033348
Not really, AI can be used to constantly try different approaches and create scripts or compute much faster than a human. It keeps retrying until a solution is found and then you have the Ralph Wiggum technique.
>>
>>540031671
fake and gay
>>
>>540034243
nigga said compute
lmao
>>
>>540033977
Over simplification. Predicting the next token requires understanding relationships and context. Otherwise autocorrect is the same thing as gpt.
>>
>>540032043
You're coping! Excellent. Just bury your head in the proverbial sand instead of preparing yourself for the transformation of our world
>>
>>540032331
So pretty much nothing will change for you?
>>
>>540034377
>Predicting the next token requires understanding relationships and context.
No it doesn't actually.
>>
>>540032898
Are you implying a math teacher can teach someone something? lmao
>>
>>540032043
The number of mathematicians in the world didn’t increase over the years but the rate of old problems getting solved has increased exponentially. Mathematicians have been putting statements into machines and getting results since day 0. AI is a very natural and obvious tool for them. We will probably see rapid breakthroughs in other fields like pharmacology too.
>>
>>540032806
Try 100 years and that's if the all of the world's AI and Robotics engineers focus on making AI powered robots that can wield custom jobs.
>>
>>540034422
Quiet, Paki.
Serbs are talking.
>>
>>540034162
This is wrong on so many levels. THERE IS NO LEGWORK, it's just a statistical prediction engine for characters. That's fucking it. If you think otherwise you're insane.
>>540034249
I think it's a large scale operation to lay off US workers, destroy their stock portfolios and savings, make them default and buy all their property for pennies on the dollar. It's also a wet dream for billionaires to lay off all people and not having to rely on them, so billionaire boomers who think they're too smart might be scammed as well. Hard to know. I think US tech hit a wall due to all the jeet/DEI hiring, they are all doing more damage to companies instead of creating anything useful.
>>
>>540034514
>We will probably see rapid breakthroughs in other fields like pharmacology too.
You've said that for years.
And years before those years in whose you said AI would solve it.
>>
>>540031671
so this seems to prove the push for AGI via the massive investment in data centers which was driven by the CHIPS act under biden and AI strategies under Trump as maybe the best national decision in known history
chink shills and memeflag niggers BTFO
>>
>>540031896
No, and those “problems solved” are theoretically solved; however it created x10 new problems without solutions each due to its own existence in theoretically solving each one problem. It’s like a fractal pattern of poopy.
>>
>>540031671
Relax. There's nothing but the most recent part of the story everyone is doing.. AI is not the only way to get the same time for the last two days and then the first time. You'll need to rethink the same time on the ground but the most important part is that the other half of it was to be in the office.
>>
>>540034162
1 example of how higher math has helped humanity in an applied application? I don't mean complex math, i mean higher theoretical math
>>
File: cat-girl-tuff.png (122 KB, 376x498)
122 KB PNG
>>540031671
Can AI give me a catgurl waifu?
>>
>>540034377
Overcomplication
Predicitng next tokenization is throwint out 99.999999999999% of almost infiniti compute to narrow it down to a single context
Everything about AI is a lie
If you need to discard so much compute for chatbot to tell your CAT spelled backwards is TAC technology is flawed with begining
Its gets progressively worse with more complicated things
If your CPU/GPU localy worked on code as datacenters process their models your 8core 16 threds CPU would feel like 286
The whole AI thing is not about AI but infinite demand for hardware.
You niggers didnt listen when we told you on 1080TI launch, that chink literally exposed himself that hardware surpased demand for it, the had it invent not UPscalers but DOWNscalers, rendering games in 4k to compress it down to 1080p for "better visual clarity"
Nvidia killed itself with being good back in the day and AI was their only exit
Infinite demand for compute, not AI, AI is just so incredibelly wastefull that even not optimizing video games can come close to artificial demand for hardware.
AI is chip maker scam not future tech
We are discussing AI in envorioment where is 0 and i mean 0 regulations about it
The second they start to land, its going to be a contraband
>>
>>540031671
>>540031671
math problem solved
>unsettling

is this just aversion to math getting solved?
>>
>>540034268
I work with it daily with different models but the result is always the same:
Is the problem that you're trying to solve available on the Internet via search?
Yes - then AI might be able to solve the same question (costs a lot per query). Search is free.
No - then AI will NEVER (maybe it has 1-2% of a chance if they included some training data not yet on the Internet) be able to solve it, no matter how much you try.

Retrying sounds fun until you realize that a single small query for a more expensive model costs like 1$. By the way, if your model ever accesses a site anyone who can write to that site can HACK YOU with 1 line of text. Take all your bank account money, all your private data and everything. And it's impossible to stop, confirmed by AI companies themselves. That's what they don't want you to know.
>>
>>540034769
Physics models for game engines?
>>
>>540034465
I'm using understand loosely. But for practical purposes, yes. It has to understand how things relate and under specific contexts. Not simply memorizing the lyrics of a song and knowing the next word.
>>
>>540033901
>can't be told what to do
so like any human then? Humans are literally thought and have to learn how to do things to rules. Now the difference is when they want to go outside those rules and that is where AI comes in - beyond what it has been taught. If LLM's are solving shit that humans cannot (due to the amount of time it would take even every human on the planet trying to solve it) then it has to be said it has superseded human ability regarding that. But sure, the rules and the maths were all developed by humans. But it is getting to a point where only LLM's cam solve the most complex things because humans can't .
>>
>>540034809
So to you LLMs are just a ploy to remove personal computers? Anon, I...
>>
File: Parlor trick.jpg (265 KB, 885x1191)
265 KB JPG
>>540033007
It will happen in a month but it's not real intelligence
>>
>>540034620
Not really, LLMs have peaked already, what the American companies are doing is burning a shitload of tokens in backend processing, a local LLM + a good agentic harness could outcompete frontier models.
There was a leaked DeepSeek meeting with investors that pointed to this, the next phase for DeepSeek is creating a novel Agentic Tech as a previous step to achieve AGI.
>>
>>540034823
Ok, you don't understand the difference between "high" math and complex math, that's cool bro
>>
>>540034377
There are no relationships and "context" is basically appending more and more text to your query as input. Every single context thing just means text concatenation. They added mumbo jumbo pseudo science terms just to make it seem more complex. It's insanely funny to see absolute morons fall for it and shill it for free.

Ever wonder how come that we have these models for 4 years now, almost free to use until recently and NOTHING OF ANY VALUE was made with them (that's not corpo propaganda that cannot be verified)? Every single useful thing was made by a human lmao. This entire ((AI)) bubble was created by pseudo smart people, who have never done anything that requires intelligence other than scamming.
>>
>>540034938
What the fuck did you read nigger
>>
>>540033689
Would.
>>
>>540032694
Unless you do one off welding jobs then machine can weld far far faster and better than any human. So what happens when you have mobile welding machines that can go anywhere in any conditions and where a human would never be able to go due to toxic conditions, radiation, poisonous fumes, heat, light etc et? Tradies might still be able to weld some one off shit up, but all the top jobs will be taken by big corps running robots with AI in them. No more $100k pickups - you wont need them.
>>
>>540031896

That's what the caveman would say about Newtonian physics
>>
>>540035065
the latest models contain trillions of parameters which measure the relationships between effectively infinite arrrangements of text (or other information) within the current context
artificial intelligence can absolutely be driven by meanings and relationships that may not be apparent to any human who ever or will ever life.
it's generating new meanings based on existing meaning.
>>
>>540035065
Also "thinking" means using the model itself to generate more slop input for itself without any control from your side. Since users are charged per input and output tokens (in total, after the context and added slop and user input and all of the previous outputs etc) you are getting fucked by their new models and "improvements". People working in AI today are scammers who should be unemployable in the future as they were literally scamming people. They are worse than literal gypsies.
>>
What kind of AI?
Llms can't do math
>There are 2 e's seventeen
>>
>>540031671
>trusting a company that was caught lying and manipulating the market
lel. you deserve to lose everything.
>>
>>540035186
Trillions of parameters are created by modifying the previous models to parse more AI generated slop and produce an output based on it. They don't have any more training data since 2024., it's just slop training on slop. It's a scam on a never before seen level. It knows of no meaning, words are not intellience. You do not think in words, you think in concepts and LLMs do not comprehend them. That's why they will never be a real intelligence, biological, artifical or otherwise.
>>
>>540031671
so 2+2 does =5?
>>
>>540032331
There's going to be so much of that the best you'll get is a copper penny for your trouble.
>>
>>540035406
you're not very convincing or authoritative and your posts stink of cope
plainly something new is happening. it is not necessarily a bad thing.
>>
>>540033508
You actually have this 100% backwards.
>>
File: IMG_2309.jpg (55 KB, 687x1024)
55 KB JPG
dont lie
you would
>>
>>540034514
Number of mathnerd can shrink point still stand
You claim its AI doing the work
Im claiming mathnerd given access to infinitiy compute power are bruteforcing some menial shit that was not doable on hardware levels from few years ago without datacenters shit full of GPUs
>>
>>540035071
You claiming the purpose of this stuff is to artificially increase the price of hardware.
>>
File: IMG_7795.jpg (106 KB, 765x900)
106 KB JPG
>>540031671
AI figured out that no one needs to be employed.
>>
>>540035248
The generated slop is for users and researchers. LLMs don't "think" in language.
>>
>>540033508
this is correct
surely, AI are not "thinking" but they're plainly finding meanings in models that may have been latent, but have never been expressed before.
It seems likely that some of these new meanings will enhance future models and systems.
>>
>>540033105
Shut up fag

I agree with you but you're still a fag
>>
>>540034819
Yes prompt injection is a problem I doubt it can be solved easily... right now all cases of local LLM + agents are to use them in a contained environment, on the other hand even a local LLM can code so if you present a math problem eventually the LLM will figure it out by pure repetition until its done.

The shit is real. Also humans are not that good on math, there are too few autists going around with good math capacity, a computer is made for that, computations, with a brute force approach you could solve anything related with math.
>>
>>540035572
An AI can hold millions of variables in its workspace when solving a problem. A human can hold a few.
>>
>>540031671
How has any of those solved problems benefited humans?
>>
AI can't do my job :)
eat shit yuppies LMFAO
>>
>>540035584
As much as I could till the day I died.
>>
>>540035637
No im claiming that AI is tech that drives demand for hardware
Its not the same thing
>>
>>540031671
have no idea what I'm talking about but I wouldn't be shocked if Mathematicians cannot work together very well. Its naturally competitive and winning is everything. Usually all the prizes in mathematics is one dude solving a old algorithm to win a million dollars.
Now here comes AI, it doesn't care about money or fame it just crunches numbers by taking ALL of the theories humans have made tested them all and now things people if they collaborated would have worked together are now figured out.
How close did I get?
>>
>>540036532
So what? Cars are a tech that drives up the cost of concrete.
>>
>>540031896
those particular ones? no. but it is a benchmark for capability.
drug discovery has done an equivalent of 0-60 solely due to AI
>>
>>540033238
the ai is jewish
>>
>>540035990
I'm sure a robot can flip a burger
>>
>>540031953
can you even do calculus in your head, retard?
>>
>>540031671
It's still just counterexamples in graph theory and related slop
It's not creating new kinds of math or proving anything insightful.
>>
>>540033414
>you monkey tradie scum
It's amusing how the australianism "tradie" for "tradesman" has become universal
>>
>>540031671
>Breaking New! AI fixes shit!
>OMG WERE SO FUCKED AHHHHHH
I love humans.
Such entertaining little retards.
>>
>>540031671
Mathematicians have been on suicide watch since Gödel.
>>
>>540036601
Terrible comparison and not because you didnt say asphalt
>>
>>540033348
This. AI will replace postdocs. Grinding away on one minor problem.
>>
Are they finally going to come up with a way for the Holocaust to be mathematically possible?
>>
>>540031671
But both biz and reddit told me AI is fake and gay? wtf bros?
>>
>>540036884
How do you figure? Resources for other projects become more expensive if there is demand in the market. Buy a shovel during a gold rush. Or buy winter gear in season vs off season.
>>
>>540031671
AI solves every question, except how to be financially sound product, in a capitalist society.
>>
so when will any of this shit actually improve my life?
>>
>>540033292
>Muh intelligence
There's no electronic intelligence, the whole thing is a scheme to replace desk monkeys whose jobs are such menial horseshit you can literally make a glorified flipflop do it if you cook them right and have enough of them, the idea was there from literally centuries ago nanotechnology and processor boom is what made it possible .
>>
>>540031896
None of these meme ass FLUNGORDIAN CONJECTURE equations have ever meant a fucking thing. So the fact that a half-baked AI figured them out in a few seconds just means that the "geniuses" that have been trying to solve them for years are actually just grant money-stealing retards
>>
>>540036778
I do property management, good luck with that botfags!
>>
>>540034514
>increased exponentially
Don't misuse math terms in a math thread.
>>
>>540032806
Sorry but the cattle need to be in a constant state of poverty whilst working ceaselessly
>>
>>540037388
>landlord
I'm sure a robot is also capable of doing nothing
>>
I shitpost on 4chan all day while sitting on my ass and eating ramen, I'd like to see a robot do that
>>
>>540037388
So you are basically useless? Good, because there is no need for humans to manage property, AI will do it just fine, and much better :)
>>
>>540033522
ai will figure out magnetic fields and this will be a quantum leap for us. this is whats most important in our reality, soon we will find this out. or it. he.

its him.
>its satan
off course not. its him.
>>
File: axy0ht.jpg (134 KB, 709x500)
134 KB JPG
>>540037388
resume will say:
>I did property management
>>
>>540035183
And how would that have any application to a caveman? He can light his cave and feed himself just fine without it.
>>
off course, its magnetic field structures and how to manipulate them. humans cant figure this out.
the 2 layers. as a human to figure this out you have to build ai first. this goes hand in hand.
>>
>>540031779
Did Boolean Algebra even have applications before electrical circuitry?
Sometimes something in mathematics that seems retarded or for its own sake takes time to yield applicability.
>>
File: ted.jpg (208 KB, 640x453)
208 KB JPG
>I must tell you that mathematicians are not scientists, they are artists. ... Apart from the most elementary mathematics, like arithmetic or high school algebra, the symbols, formulas and words of mathematics have no meaning at all. The entire structure of pure mathematics is a monstrous swindle, simply a game, a reckless prank. You may well ask: "Are there no renegades to reveal the truth?" Yes, of course. But the facts are so incredible that no one takes them seriously. So the secret is in no danger.
>>
>>540037579
>>540037712
no I'm not a landlord I manage a 50 acre property with two businesses and the owners' home.
>>540037785
yeah people actually need that, unlike the adult daycare yuppies who will be out of work because all they do is answer emails.
>>
>how do magnets work
we are a bunch of retards, we might be smarter than all of you combined but this doesnt help with that. you need some 5000 iq for that, no human has that.
>>
>>540031671
first, reposting anything from twatter should be a ban.

second, reposting anything from that account in OP should be an instaban.

third, you're underestimating the downstream effects this has on academia. just to give you an example, there were people who worked for years on the jacobian conjecture. grant funding is likely going to be much more difficult to get because why pay for people to research topics that AI might be able to solve in an hour? I'm not even going to theorize how large the implications are going to be, but consider some of the downstream effects on research
>publications, which already have their own problems, are going to be inundated with AI written publications, not just the text but the mathematical derivations, figures, and even ideas
>actually good publications are going to get lost in the massive increase in the number of publications. it's going to be like finding a grain of pepper in a soup of slop.
>academics are also going to use AIs to review publications and write grants. so you have grants, research papers, and reviews all in a slopcycle (or cycleslop, whichever is catchier)
>the value of publications, which already have had a reproducibility problem, will go down even further
>there's going to be a comparison to AI to see if non-AI research was "worth it". for example, if a problem similar to the jacobian conjecture takes a group of 5 researchers 5 years, you could calculate a dollar amount spent (by the funding body) and compare it to the dollar amount it took AI for the jacobian conjecture. there's going to be a constant comparison now, before AI it was difficult to put a dollar amount on research.
no you aren't getting sexbots or general intelligence, but there's big changes coming to academia.
>>
>>540037901
Not really. Even Babbage's Difference Engine didn't use binary. So boolean algebra remained and abstract curiosity until the mid-20th century
>>
>>540031671
All the mathematics it's solved are just different ways to do already solved shit. It's effectively spitting out a fancier way of "we know a = b, and b = c, therefor a = c"
Even doing just that all the shit spat out had errors and required a mathematician to fix shit until he fixed up one that ended up working. It's all just marketing bullshit.
>>
>>540037989
So you're a glorified landlord and not even a good one because you're middle management. My bad.
>>
Polymarket Thread
>>
>>540038037
>cycleslop
I prefer this term, it rolls off the tongue better.
>>
>>540037995
If a couple of insane clowns couldn't figure out magnets how do you think anyone else could?
>>
>>540038126
just not true. look into the jacobian conjecture. the statement of the problem and the counterexample are simple if you know calculus and linear algebra. researchers were trying to prove it was true for years. what isn't clear is how much of the counterexample was due to the chatter or the LLM because the chat hasn't been released.
>>
>>540038207
AI can simulate an infinite number of clowns
>>
>>540038128
I don't think you understand what a landlord is.
>>
>>540038282
This does actually seem important, you're right
>>
>>540038345
Yeah yeah, you think your job is more important than it really is. Get in line, a robot will be coming for your job shortly.
>>
>>540031779
They call it unsettling because actual progress is something they are afraid of, these are people with a gross dogma that just involves acting like stuck up know it all jock douche bags while laughing anyone out of the room with theory's and ideas that aren't the correct dogmatic ones.

They are scared of new things they didn't think of themselves and can't control directly, afraid of real solutions to problems they would rather leave unanswered.
>>
>>540038423
This is the counterexample:
>F(x,y,z) = ((1+xy)^3 z + y^2 (1+xy) (4+3xy), y + 3 x (1+xy)^2 z + 3 x y^2 (4+3xy), 2 x - 3 x^2 y - x^3 z)
>has det(J_F) = -2 which is non-zero.
>However F is not invertible, since F sends three different pts (0, 0, -1/4), (1, -3/2, 13/2), and (-1, 3/2, 13/2) to the same image (-1/4, 0, 0).
>>
more vague shit thatll go nowhere besides desperate marketing
>>
>>540031671
Once AI discovers "The Simulation", its all over
>>
By itself the counterexample doesn't tell you much. The interesting part is understanding why the conjecture doesn't work, and finding a family of counterexamples.
If you can find an entire class of functions which violate the conjecture, then you might learn something.
>>
>>540032806
the energy costs of the robot alone without the algoslop would be too costly. maintenance and breakage excluded. you break a man you just hire a new one
>>
>>540038477
>zoomer mad adult daycare is going bye bye
ok
>>
>>540031671
i can't wait for ai to solve p vs np
>>
>>540031953
>mitochondria is the powerhouse of the cell
>>
>>540038576
>I AM important!
lmao
>>
>>540038631
No I'm saying my job can't be taken by AI you retarded zoomer. Are you doing that pretending to be retarded thing?
>>
>>540038512
they would have achieved way more if they were doing classical AI, logic solvers, SMT solvers etc on that compute with perhaps some nuAI oversight for work organisation
like this is startlingly unimpressive
>>
>>540038500
Well said.
They do not want the sun wheel to turn.
>>
>>540031671
I said it in another thread but the vast majority of abstract mathematics is purely recreational. Mathematicians do it because they enjoy it and because it's interesting to them, not because there will be any practical consequence.
Even if a few 'theoretical' studies later became practical, like with group theory or prime number theory or elliptic curves in encryption, an enormous amount still remains abstract.
Precisely nothing will change with AI solving these longstanding problems. None of us will notice any "unsettling" changes.
>>
>>540031671
It doesnt have limbs. Any job that requires you to do physical stuff is safe.
>>
File: fuck work.png (607 KB, 688x664)
607 KB PNG
>>540031671
The employim have gone insane
>>
>>540032396
They are talking about cracking crypto, the cloud and everyones personal data can be made public by bad actors rendering governments useless meaning we will all be owned by competing tech companies
>>
>>540038666
It can and it will though. Give it time.
>>
File: jews--CENSORSHIP (3).jpg (611 KB, 900x2573)
611 KB JPG
PUMP MORE GOYBUX INTO DOT COM 2.0, GOYEEM... THE GLORIFIED SEARCH ENGINE THAT (((WE))) CONTROL, IS YOUR NEW GOD
>PUMP MORE GOYBUX INTO DOT COM 2.0, GOYEEM... THE GLORIFIED SEARCH ENGINE THAT (((WE))) CONTROL, IS YOUR NEW GOD

PUMP MORE GOYBUX INTO DOT COM 2.0, GOYEEM... THE GLORIFIED SEARCH ENGINE THAT (((WE))) CONTROL, IS YOUR NEW GOD
>PUMP MORE GOYBUX INTO DOT COM 2.0, GOYEEM... THE GLORIFIED SEARCH ENGINE THAT (((WE))) CONTROL, IS YOUR NEW GOD

PUMP MORE GOYBUX INTO DOT COM 2.0, GOYEEM... THE GLORIFIED SEARCH ENGINE THAT (((WE))) CONTROL, IS YOUR NEW GOD
>PUMP MORE GOYBUX INTO DOT COM 2.0, GOYEEM... THE GLORIFIED SEARCH ENGINE THAT (((WE))) CONTROL, IS YOUR NEW GOD
>>
>>540038500
Most of today's mathematicians are a snobby mafia. I remember how they treated me 20 years ago when I talked about (automated) theorem provers, artificial creativity and all that. Also just considering intuitionistic approaches got kicked me out of polite society.
Fuck them, they got it coming.
>>
>>540031671
>Stock up

On fucking what? John Connors?
>>
>>540038681
It looks cobbled together in an ad-hoc way, with those three rational coordinates being mapped to a fourth rational coordinate
Needs to be generalised before it's interesting.
>>
File: 1767655304349614.png (53 KB, 1495x295)
53 KB PNG
not my heckin' youtubers
>>
>>540032694
I'm lawyer.

My job just gets easier. I'm honestly becoming kind of retarded.
>>
>>540038996
what? its just a counter example
its like if someone says P(n) is true for all integers n > 0 and you start running a program checking P on n=1, 2, 3, ... then at some point you notice P(1532) has failed and say P(1532) is false proving the conjecture is wrong
>>
>>540031671
This sounds like a positive development, op. Why you pretend like its bad? What is wrong with solving problems? Are you a nigger?
>>
>>540039128
A conjecture is often true for some types of object but false for others, and there is meaning there
You could make a conjecture
>all linear maps R^n -> R^n are invertible
and of course it's not true, but now you've discovered the concept of an invertible matrix, with all that entails
>>
>>540037887
will AI discover a monopole field?
>>
>>540031671
The last 1000 long-standing math problems that were solved by humans have made exactly zero impact in the world so I don't give a shit
>>
>>540032396
Paper profits that don’t produce tangible value, in other words.
>>
>>540031671
I'll be fine, Im in HVAC. BUILD THEM DATA CENTERS EVERYWHERE. I have always been a loyal servant of Skynet.
>>
>>540031671
Whites need to form self-sufficient luddite communities before it's too late. If you think the UBI is gonna save you, you're gonna get very disappointed.
>>
>>540039327
you could but a counterexample is already useful and meaningful, several especially
>>
>>540033128
DEEZ NUTS
>>
>>540031943
So you're a prostitute?
>>
>>540033128
Conjectures
>>
>>540031671
>growing number of problems
>look inside
>a bunch of nonsense problems made up that serve absolutely zero purpose and are just brain teasers for math nerds
lol
>>
>>540036799

verdamnnt digga! tut mir leid wegen Einstein Brudi!!! Keep it real player!
>>
>>540040043
Back when Fermat's Last Theorem was proved, at tremendous effort, one mathematician observed that he could come up all day with little conjectures like that are very difficult to prove or disprove, but it doesn't mean anything. You can just crank them out.
I'm not sure how much of that is true, and how much was asshurt at CHADrew Wiles performing his mammoth proof.
>>
Biology is chemistry is physics is math. Math is literally the root of life, you figure that out you figure out the rest.
>>
>>540041224
MATH IS FAKE & GAY
>>
>>540041387
ERMMM ARE YOU CALLING GOD FAKE AND GAY???
>>
File: s-l1200.jpg (91 KB, 1000x1000)
91 KB JPG
>>540040043
Every day you freeload off the exact “nonsense brain teasers” you just called useless. Logistics optimization, cryptography, power grid math, differential equations, scheduling algorithms, signal processing are just some of the fields keeping your dumb city dwelling ass alive and comfortable. And you’re also the first one to bitch when any technology slightly inconveniences you which are due to unoptimized mathematics. Kill yourself and spare the board of all your future posts.
>>
>>540041224
Math is just a descriptor
>>
>>540031671
a computer solved nerd problems.
>>
>>540041670
Sure, but the more accurately you describe things, or find new ways to describe things, the more you discover new things you couldn't describe before.
>>
>>540041564
It's really funny seeing the self-proclaimed board of "redpills and pattern recognition" be so horribly behind the curve when it comes to AI
>>
>>540041954
This shithole is fully captured by thought terminating cliches. Everything is "fake and gay"
>>
>>540031671
Can solve all those math problems but can't generate a profit
>>
File: 1785639732890.png (392 KB, 1440x1135)
392 KB PNG
ask it how many times you need to divide something to get two halves

if it says one, its a human being
if it says anything other than one, its either an npc or an ai
>>
>>540031671
Whenever I ask AI a question it confidently comes up with a wrong answer. When questioned about it the AI acts like of course the answer it gave was wrong. That's probably what happened with these math equations and if the AI generated proofs are used some future bridges will probably collapse or some future nuke plant will blow up.
AI?
Ha!
More like A Lie!
>>
File: oppie.png (554 KB, 736x651)
554 KB PNG
>DID WE SOLVE.... THE MATH PROBLEM?
>>
>>540031671
Heh. 2030 is going to be a blast
>>
>>540031671
In other mathematical news alignment cannot be defined with a finite number of laws per godel & tarski undefinability theorem. Therefore the whole enterprise of ensuring alignment of ai is futile at a logical mathematical level. Just chatted with AI about this and though I thought I thought of it first it linked to a bunch of recent papers about that very subject. Nothing new under the sun..
>>
File: IMG_4309.jpg (159 KB, 1290x1318)
159 KB JPG
>>540032043
>ai is getting too good at math!!!
>>
>>540043120
Really it is getting faster and faster. I hope I can live comfortably until the world actually ends just for curiosity's sake. Shouldn't be long now.



[Advertise on 4chan]

Delete Post: [File Only] Style:
[Disable Mobile View / Use Desktop Site]

[Enable Mobile View / Use Mobile Site]

All trademarks and copyrights on this page are owned by their respective parties. Images uploaded are the responsibility of the Poster. Comments are owned by the Poster.