How is AI going to change your country? Demis Hassabis (super genius), just moved his AGI prediction from 5 to 2-3 years. It's insane how much better it got in just the last 3 months.
>>225855375Gemini is fucking useless and chatgpt s shir iclaudes the only decent one
What happens to India in an AI age?
>>225855488>Gemini is fucking uselessmaybe for your programming monkey jobi have a real office job and Gemini is excellent, especially with reading photos of random pages and stuff.
A frontier model (currently not out to the public), created a novel physics problem nobody has previously thought about, solved it, wrote a paper, and published it to a top journal. And did it 3 hours.
>>225855569mathcucks and physicians at academia on suicide watch rn
AI devlopment has already hit diminishing returns. It will not be significantly more advanced in 3 years.
>>225855801not true
>>225855375I've already got replaced by artificial intelligence, so I do some delivery jobs, construction works, popup booth runnings and dishwashing since I got out of the job
>>225855801People were saying this last year and look at where we are now
>>225855593Math dorks were doomed the moment a calculator came into existence
>>225855520pajeets cheering for ai is honestly hilarious if you know that india's entire development strategy relied on low value service jobs which were obliterated by ai
>>225855375They said the same thing 5 years ago, that AGI was just 5 years away.
>>225856044Jeets claim they will have a frontier level model by 2027 kek
>>225856021Basically the same as last year?
>>225856138AI wasn't solving legacy mathematics and physics problems a year ago like it is now
>>225855527gemini is useless even for that, it cant even scan a schematic for me. what kind of babby work do you do?
>>225855375Claude beat Pokemon Red using only vision, the thing is basically as sentient as a 5 year old.
>>225856177>what kind of babby work do you do?reports for the judiciary system
>>2258560211. I didn't hear anyone say ut last year.2. 2026 AI slop is barely different from 2025 AI slop.
>>225856169>implying that solving these "problems" will change anything, they're symbols on a screen, nothing will change
>>225856265Randoms making slop has nothing to do with AI
>>225856211A bot can beat you in Unreal Tournament 2004, that doesn't mean it's sentient.
>>225856138The type of shit you are able to generate now outscales what you were able to generate 2-3 years ago.
>>225856300The primary output of AI has nothing to do with AI?
>>225855375>AGIjust scaremongering to keep their stocks on top like a good businessmanAI is just a language learning model, it doesn't have initiative unless someone have some intentions with it
>>225856138in the last few months AI has started solving some really serious math problems, including Fields Medal stuff like e.g. quasi-Riemann hypothesis
>>225856314No like, just vision. Completely untrained on it. A bot can't beat a pro player in anything without training.
LLMs are generally intelligent already and smarter than most peoplewhat they mean is magical super intelligence and you can just move the goalposts for that forever
>>225856039Lol you dumb shit. Do you think math is about adding large numbers? It is about developing new mathematical concepts or frameworks. AI will never be able to do this. AGI will never happen.
>>225856211Claude was helped by her fellow AI porygon
>>225856336Most people use the internet for garbage, doesn't mean the internet is garbage.
>>225856512>AI will never be able to do this.It already has. Terrance Tao put out a video basically crying and saying mathematics is over.
I predict a two phase crash. First recent gains will suffer from expansion fatigue causing the bubble to burst. Second an Al black hole effect will arrive a few years later where Al becomes too advanced and devours the free market that originally sustained it.>>225856294Al becoming capable of doing research level math is exactly when things started to get serious. Remember it is a creature of math itself. All algorithmic problems boil down to mathematics if Al can improve those algorithms it might also improve itself. This is where the sudden fearmongering around RSI came from.
>>225856573He was crying because AI provided a functional solution, but didn't do a proof. As if that ever has mattered. Geometry has been around for like what almost 3000 years? Euclid's 5th postulate was never proven and geometry still works. He's actually throwing a tantrum and thinks that cancer patients actually would care to know why a medicine works even if they already know it works.
>>225856671>AI provided a functional solution, but didn't do a proof.But it did provide a proof. It's just too hard to read and there aren't enough mathematicians who can keep up with ever growing number of it.>Euclid's 5th postulate was never proven and geometry still worksYou don't understand how axiomatic systems work.
