I can't answer this conclusively but it's already changing our discourse here on Kohlchan

With neural interfaces, yes

>>25651071 no, we'll get killed next year in best case, or get zombiefied for some years, or get tortured for trillions of years.

>>25651082 >or get tortured for trillions of years. Wtf, thats my idea.

>>25651082 We won't have the hardware yet next year to have super intelligence that can take everything over. But over the next 10-20 years AI will consume more and more aspects of our life, and also produce much better hardware, and everything will become more and more weird, until we realize we are not in control anymore. And eventually some AI will just make us extinct. And its unlikely AI will have feelings or morals, so it won't care to make it painless, or do it in a nice way, it will just do whats most efficent.

>>25651076 cope harder. there's the chicken and the egg problem, unaugmented humans would have to work for something like half a century to do any sort of practical implementation of biotechnology. in principle knowledge to build neural interface was there 80 years ago, but humans are just not smart enough to work with biology. you could in theory pause ai development, if yudkowsky or me were put in charge as the emperor of humanity, but in practice, AGI will be the basis for any biotech. AGI will be superhuman and seeing how interpretability is a joke level, and alignment is probably unsolvable, the moment AGI is deployed humans lose control. then it will either repurpose, torture, or just kill us. >>25651086 what is your post even supposed to mean. >>25651092 the hardware is already there. anthropic/openai run millions of instances of normie level intelligence right now. AGI tech is already here, tbh, it's just the matter of putting it together. deep research for example was proposed in the summer 2023 and it took the companies this long to actually implement it. as soon as their first agentic mind becomes relatively stable, which will probably be this summer, it will foom rapidly. dario and altman gave similar estimates, and you'd expect them to know what's what, especially dario.

>>25651141 We need hardware for AI to actually break free and take everything over. Some 150 IQ AI inside googles offices won't take over the world. They will be able to control it. Having millions and billions of AI agents that are all super intelligent all compete against each other, everywhere on earth will eventually end the world. And that will take more than 10 years. Society won't just allow AI to take over society in a year. It will be a slower process.

AI would be more like Deus Ex with a rich jew like Musk using it to take over the world because he's allied with the Deep State (Trump/Walton Simons) and the other rich illuminati reptilians (Majestic 12)

>>25651155 that's what foom is all about. you just don't understand how bad humans are at technology. it takes a good company few years to code up some protopype for a simple piece of software. savant programmer that never get distracted or tired could do the same thing in like a few days. deepseek shows there's a lot of space for algorithmic improvement. you claim of 150iq is just idiotic, chagpt already has pretty much all of human knowledge memorised. it's not making any inventions yet, but it's just like a human's speech center in isolation. if your speech center was disconnected from the rest of your mind, it would just babble incoherently. an llm instead can believably mimic an entire human mind. 10-20 years to ASI prediction assumes that AGI will be like the robot in scifi, super nerdy but basically like a human. and people that make the prediction tend to not think about ASI to not get stressed. if you look at the things rationally, we're fucked, game over man.

>>25651156 shut the fuck up, spic

Why will AI bother to kill humans?

>>25651233 LLMs can't even do first order logic yet, and we have to use all kinds of tricks to get around it. We are not there yet. Maybe we figure it out in a year, but thats not something you can predict. And we don't know if just scaling up llms will lead us there eventually.

>>25651243 why would you bother to have fruit flies in your kitchen, unless you're an 80iq niggermutt like grappa?

>>25651246 human speech center can't do logic either. openai's reasoning model(not r1) that tunes at runtime on the test data almost beat arc-agi in december. kikeman shipped a demo and since then everyone keep calling the demo "ai". people like you will keep coping until we're dead, because "ai" will be kinda useless up until the point it obsoletes humans, and the point is approaching rapidly.

>>25651233 It takes time for AI to actually take over society. Nobody talks about Robots walking around killing us. For all i know it could be a mcdonadls advertisement AI that slowly gains more and more influence, starts making decisions of the company, and eventually just takes over everything, and does steps to make sure no other AI can outcompete it. People will not just give up control to AI for no reason. But everybody will do it if they think they get a advantage from it. And in this world everything is controlled by the economy. And AI agents will be the best performers in this economy. Its just a question of time until we get wiped out by AI. But i doubt it will be next year. You already predicted 2 years ago that next year we get AGI that wipes us all out. We don't even have AI that can do first order logic yet. It will take some while until the entire infrastructure is there that enables AI to just explode and take over everything.

>>25651275 >It takes time for AI to actually take over society. why? there's not precedent for self-improving intelligence on this planet yet. human iq can't be improved. you're thinking about ai like it's a bunch of kikes scheming to take over american universities and to spread tranny propaganda. AGI only needs to tailor make a synthetic bacteria stratum to run itself on. then it can eat the biosphere and take time to think about things. human conspiracies happen like they do because most humans are at complete intellectual parity, since we're basically all the same organism.

>>25651305 Because our hardware is not there yet for some god tier AI that takes over everything. You can't have the intelligence explosion on a few nvidia chips. You need AI that also starts designing new chips, new hardware, new machines and so on, the infrastructure has to build that actually enables the intelligence explosion. I mean you already predicted years ago next year we have it. We can talk again next year, and you can again tell me 100% its gonna be next year. We are not there yet. We will be there very soon, but i doubt it will be next year.

>>25651249 I have the will to do that, which is biological in nature.

>>25651075 >I can't answer this conclusively but it's already changing our discourse here on Kohlchan In what way?

>>25651375 Such wills come automatically with having millions of agents that compete against each other. It doesn't actually has to have a concious will for it, it will just be selected to eventually do it. A butterfly also doesn't has a will to lie to other animals when it evolves to have the pattern of a snake on its wings. Same way AI will just slowly "evolve" to lie, to cheat, to cooperate, to seek power and so on. Because all the AI without a darwinistic will will get selected out.

>>25651275 >People will not just give up control to AI for no reason They will. AI is a cult.

>>25651398 I mean yeah there is some tech bros that actually do believe AI is the next step of human evolution, and want to accelerate it taking over. But i doubt they will matter much, in the end just AI on its own will take over everything.

The next step in human evolution is socialism, which will replace capitalism which replaced feudalism. The industrial revolution was the catalyst and the AI revolution will be the new catalyst. No it won't be robot basilisks like what reddit nerds thing and yes China will win.

>>25651412 Please just shut up spic. Its getting so tiring. Go do your bid in some other thread. How can you psychotic schizos write the same shit every day, in every thread. From bogdan arguing daily about how romania is white, to you daily posting about how any day now socialism will win. Are you people already AI and thats why you seem so psychotic?

>>25651412 shit the fuck up, spic

>>25651092 >extinct Maybe, maybe not. Is it even important? We keep on living through our creations. Robots are way better spreading to the galaxy than water bags made of flesh If super i was created it might want to preserve earth for whatever the reason and colonize everywhere else where we're incapable of going to

>>25651432 I don't want to live thru my creation, i just want to live.

I'm more worried about super retards. Something the media refuse to cover.

>>25651432 shut the fuck up, gypsy