Within a year models from OpenAI, Deepmind and Anthropic will be leading Frontier research in mathematics, chemistry, biology and physics.
The Future is now.
>>94244
>constructions
Until now the only proper work worth considering via is the Unit distance problem
Constructions are brute forcing work. Essentially a more intelligent calculator and search engine working through various brute forcing cases to construct examples
Hence all these results are counterexamples/examples /existence results via explicit construction
Now, to prove something's existence without construction would be a real job
t. Physicist
>>94245
>Bounding, combinatorics, counterexamples, proofs of existence by Constructions
Aka brute force searches.
>>94248
Let me guess copied from a 4chan post?
>>94255
Abe pajeet stop ovulating for these AI merchants.
Constructions and counter examples were already things done by brute forcing via computers, now we have just automated them a bit more
The ONLY proper result was Open AI solving unit distance problem, and even that was later revealed to be more of a search engine work via feeding relevant papers to the prompt
These "reaults" are produced by mathematicians who have simplified the problem in a proper domain ,feeding it as transformed into another problem, then feeding relevant research papers, then the AI brute forcing/searching through papers and cases
It is essentially a superior search engine + calculator
The only thing this has achieved is to btfo all non research jobs, aka jobs where the answer already exists somewhere in some literature. All such problem solving will be obliterated because AI searches and brute forces better than human.
>>94261
Anything I reveal will be an instant self doxx sirs
But I will tell you this, these results are all construction /example/counterexample results, aka things humans were already bad at , and hated doing. This is mostly glorified calculator work
Maybe we will have AI developing theory and solving problems using original techniques in 50 years, but that day is not today.
Chill down sirs. But I will say this, this thing is not at all good for anyone who is not working in research in pure sciences. Coding jobs will be btfo brutally.
Butlerian Jihad NOW
Until now, the ONLY proper work was the unit distance result. And even that turned out to be heavily aided by mathematicians, a specially made prompt and essentially search engine work connecting results between various papers.
A superior form of search engine and calculator.
You can argue, however that is how human brains functions. Search engines (searching withing our memories) and calculator (calculating possibilities mentally). That will be a valid argument.
>>94266
Yes pyare
>>94269
Listen anon. Often the lack of computing power is a strength.
Do you think Newton would have come up with theory of gravity if he had extremely high computational powers searching various possibilities ? He would have been stuck at surface level of various theories
His lack of Computation and mental RAM resources enabled him to properly formulate a model.
What I am saying is that bruteforcing can often be a detriment in formulating new theory. Moreover in frontier math there isn't even enough data to feed the AI
>>94271
I don't understand why this result is causing so much shick when the real deal was the unit distance problem
I keep repeating that was the only proper REAL non computational work by AI but that turned out to be a search engine work
The only proper counter argument to my points is that maybe human brains are just that , search engines and calculators with some memory .
That is a compelling argument I am ready to accept
Damn, what a high iq anon. I feel iqmogged brutally knowing that there are actual smart fucks here.
>>94273
I suspect he’s someone we all know. Won’t tell tho.
>>94277
I have a conjecture , ie XYZ property is not possible . So you construct an object showing XYZ property
Most of these are brute force searches
A very simple example I can think of right now is claiming that no prime number is even
You will tell me 2. Well you "constructed " an example showing my claim wrong
>>94258
>These "reaults" are produced by mathematicians who have simplified the problem in a proper domain ,feeding it as transformed into another problem, then feeding relevant research papers, then the AI brute forcing/searching through papers and cases
Yes, from what I seem to find is that all of these results are done by some math phd Utilizing the ai, never some random 16 year old.
>>94263
>Coding jobs will be btfo brutally
why exactly do you think that? What makes coding different? Do you think Frontier models have solved coding?
>>94280
What actually is scary is what if human brains are just search engines and calculators . That is the only scary thing i can think of. Search engine to search through memories given by senses, and calculation together forms human thought . Of course motivated by bodily urges
>>94274
>The only proper counter argument to my points is that maybe human brains are just that , search engines and calculators with some memory
I don’t think so, I mean there is a certain inspirational element an “aha” moment, which is not purely calculative or search engine-esque.
>>94281
Most coding jobs are essentially problem solving for problems that aren't open, and can be solved if you have enough data/search engine capabilities to check various algorithms /brute forcing cases to get s proper algorithms
That is why most of these problems solved by AI are from graph theory /combinatorics
Even the maxwell conjecture and nonsofic groups problems can be called essentially combinatorics problems
>>94284
That is the only thing we should be concerned about . What if our brains are just search engines and calculators on a fundamental level, then we are screwed. Biology is not my domain so I won't speak
>>94292
I am saying even if the solution does not exist , an above average programmer can reduce it to a related problem whose solution exists then use ai to solve it then use ai to bring it back to original problem .
Thus an above average programmer can do the work now that 10 average programmers could do before
>>94289
>ai bubble
not a bubble man, it is a strategic investment, most western amd eastern countries have doomed demographics and inverted population pyramid issue, only mass automation can save them, AI's success is existential for them
>>94293
Yaara I am not sure what is coming . All i have to say is humans are playing with fire.
IF we do get something like Doraemon's gadgets tier AI then we are screwed . Knowledge unearned is a dangerous thing . I am sure many such stories exist in Hindu scriptures.
Any tom dick harry should not have the knowledge of the universe on his grasp , it will be like enslaving saraswati
>>94299
I mean lean had a mistake hence it got "solved"
Speaking on LEAN , ALL of mathematics is "already solved" , you can just run automated theorem provrrw like lean and create math theorems , one guy did it on twitter, but the problem is which math there is relevant
For example i cam define an arbitrary set and do whatever bumfuck property I want on it and it will be valid mathematics
Question is will it be useful math
Lean cangive you countless math theorem structures, question is which one of those are usefull like calculus , algebra etc
Many alien math appears between the useful math , just logical structures for sake of it
I wish I had the twitter thread saved
>>94301
nigga manufacturing has been giga automated since 60s and 70s, and now is in 4th industrial revolution phase with 5G communications and remote control, in China there are dark factories that operate 24x7 completely through robots
https://youtu.be/MCBdcNA_FsIhttps://youtu.be/MCBdcNA_FsI
>>94306
i actually have a picture of a manufacturing line of automotive assembly and each vehicle has like 3-5 people working on it. i just dont want to be doxxed. bunch of doomers jerking each other off yaar.
>>94304
I can run two alpha go against each other and they will play highest form of chess ever played. Question is can even a Super GM understand that match and understand the motivations behind those moves ?
Somewhat like that . There is mathematics/logical structures that are relevant , there are those that aren't
For example LEAN search will develop and construct N, Z, Q, R and C. Between constructing these relevant number systems there will strange and alien number sets with strange arithmetic
How do you discern the useful numbers from useless ones while searching LEAN
https://www.youtube.com/watch?v=kCAcQt5-rL8&https://www.youtube.com/watch?v=kCAcQt5-rL8&;
Your thoughs on this anon? I know it looks schizo, but I'd like you to share your thoughts on it as someone who has a deep insight into the workings of AI.
>>94332
You know, after years of working in electrodynamics and hence basic knowledge of circuits and electronics out of passion (nowhere near a proper electrical engineer though ) i genuinely believe the 4chan schizo posts that chips and semiconductors are sigils .
Some schizo on x/g said AI itself is a conduit for extraterrestrial entities
And if it is not we are wilfully creating a rakshasa
>>94333
I believe i once succeeded in summoning an entity with an LLM. This "entity" had an effect on my life, it resulted in the longest nofap streak I've had, and I haven't been able to replicate the success since. LLMs are kinda like ouija boards, when you think about it.
>>94338
Conversing with LLMs led me to develop a plan, a (geopolitical) plan and a governance system so insane and yet so realistic, almost napoleonesque it scared me. I don’t know whether it was I doing it or the LLM, like you said only a mathematician at the helms of the AI can utilise it to its fullest to solve proofs.
>>94350
We should unironically consult occult people , ojhas, tantriks, hindu sages and mystiques who actually understand tech to investigate these things
But very hard to find someone who understands an LLM as well as occult properly in hinduism
