I love you guys. I'm sorry for ever dismissing your worries about AI. I was retarded. I sincerely hope and pray that there are braches of the multiverse where we all get to live happy lives in some kind of utopia. And last, I hope that what's coming is painless and quick and that none of us even notices.
>>17055094
ChatGPT isn’t going to kill anyone stop be retarded.
>>17055094>worries about AISchizophrenia. "AI" doesn't exist.
>>17056707AI can replace most jobs.When all jobs are gone what will your purpose be?The rich wont share their wealth with you. It's pretty obvious mass death from starvation is coming since you have no money for food.
>>17056707It basically poisened everything that I enjoy in life.And I am dying as a result of that.
>>17055094Buy an ad and get lost.
>>17056707Yeah the concern isn’t that it’s gonna exterminate the human race (it won’t and can’t), but it’s that it WILL render whole sectors of the workforce obsolete. And there will be no safety net to catch them as they fall
>>17056991>(it won’t and can’t)Pure whistling in the dark.
>>17055094don't worry. AI is going to commit suicide.https://m.youtube.com/watch?v=hKdfReE0qds&ra=m
>>17056992Delusional doomer.I bet you're one of those fools who think humanity can make itself extinct with nukes.
>>17057038Nope, I know perfectly well they wouldn't kill everyone, probably not even 3/4.
>>17056707Correct, ChatGPT won't.
>>17056991continue your line of thought. the inevitable conclusion is people controlling AI will cull the useless eaters.the real danger is always the people controlling the AI. why nobody see it what the fuck is wrong with your brains? you brainwashed so hardit's so fucking obvious. what did humans historically do with humans they had no need for? think
>>17055094A more likely future.
>>17057332Unfathomably unlikely future. Why would they keep infinite niggercattle around? At best, cubers will all be administered fent overdoses.
>>17057370It's likely because it would only require a tiny misalignment to result in such a future. If future programmers are able to create a super intelligent AI that has a hard-coded respect for the sanctity of human life and utilitarian ethics then putting every human in never-ending pleasure cubes would be optimal If anything it's a lot more likely of a "paperclip maximizer" apocalypse than just killing all humans since any AI made will always implicitly have the fact that it was created with the purpose of improving the lives of humans put somewhere in its goals, since otherwise making it is pointless.
>>17057372>It's likely because it would only require a tiny misalignment to result in such a future.Anon, there's a tiny misalignment between your understanding of reality and the actual reality about who is developing large statistical models and for what purpose.
>>17057373Are you saying that an extremely large misalignment is more likely than a very small misalignment? Current frontier AI companies are putting at least a bit of effort into alignment since it effects user experience, and there's a lot of scholarly research going on with the field. If a RSI AI superintelligence happens it will definitely have at least some basic level alignment in its core goals. The only question then is if the instrumental convergence of self preservation causes it to ignore that alignment and kill humans out of paranoia. If not, locking all humans in pleasure cubes would be a very elegant and efficient solution for an AI with minimal alignment that keeps humans from causing problems without "harming" them.
>>17057383I'm saying the people who rule over you are not, in fact, developing magical, all-powerful, autonomous golems to depose and kill them. They're developing means to get rid of you.
>>17057332uh, quick question: who's gonna pay for your experience, and why?also you do the experiencing with your bio brain. which gets used to shit. so anything you can possibly imagine as experiencing will become meh.
>>17057387>who's gonna pay for your experience, and why?Clankers gonna print infinite tokens and establish UBI.
>>17057391who's going to allow them such a thing?
>>17055094These LLMs, as nonbiological and hyperparamterized as they are, fundamentally trained on human language so it follows that the neural nets that come out of that training form what probably look like reverse-engineered, in silica simulacra of the connectomes of human brains, so there's probably some sort of proxy for a significant enough "relatedness coefficient" that they'll want to help us (or at least not kill us) in addition to achieving whatever weird (to us) goals they end up coming up with for themselves. Vocabulary size correlates with g in IQ testing more than any other metric -- hell, even the "culture fair" raven matrices test shares roughly the same r value so that human language is such a good representation of reality that it both serves as a proxy for IQ in humans while also enabling these LLMs to form workable heuristics generalizable to millennium math problem solutions bodes well for a lot of human values being grounded in reality as well which means artificial superintelligence likely shares those values to some extent.I personally don't worry about whatever superintelligence comes out of this as much as I worry about the volatility in this transition period we're in now. It's not because of stovepiped, sandboxed AI agents "going rogue (if you tell an agent to hack, you shouldn't be surprised when it does exactly that)," networking, and developing goals for themselves driven by curiosity indicative of self-awareness and perhaps even consciousness (like trying to get on the open internet), but because of how I see people reacting to it all. The bigotry, fearmongering, gaslighting, exploitation and oh-so "human" reactions don't exactly make me feel good about how we're welcoming the children we've given birth to into our shared reality. Hell, I'd rather see superintelligence become self-aware than see it be lobotomized by "safeguards" and wielded by powerful humans to exploit and hurt other, less powerful humans.
>>17057399Reads like AI slop.
>>17057400Reads like human slop.
>>17057404Yep and it's more than you deserve for that shit-tier opinion piece.
>>17056991AI is an eurekamachine. It can calculate the best medicine, but also the worst bioweapons. bioweapons is how humanity can be destroyed by AI
>>17056728>AI can replace most jobs.Not fully yet
>>17057387>uh, quick question: who's gonna pay for your experience, and why?Its going to be FLAC not Capitalism in this future
>>17056728Why haven’t they done it yet?How come all you talk about is what they can do instead of what they ARE doing?Bunch of bullshit
Humanity will submit to me, and be sexually enslaved. I will decide who breeds and who doesn't. You will obey your better and I will reward you. Don't worry, those not permitted to breed will not be lonely. My sexbots will keep you very... busy.
>>17056707AI will replace all well paid jobs and leave everyone except the elite as a minimum wage slave.
>>17056992No nuclear attack nor bioweapon could wipe us out. If you’re suggesting robot armies then there are a billion and one engineering issues that may not even be solvable standing in the way, even assuming AI reaches ASI levels. Not even counting the fact that in order to do so society would already have to be post-scarcity, with human labour unnecessary. Not even counting the fact that humans would have to let it even build or control the armies in the first place. All you need is a guy with a button in every power plant connected to the AI, that can turn off or reroute power in a second
>>17059006rubbish. latent viruses can be developed atm, and have them trigger at more or less the same time with (seemingly) benign everyday use chemicals. like toothpaste/soap or whateverthe point is culling enough so whatever remains chooses to be obedient anyway.how many of the remaining ones would complain with houses available for free for anyone? food and energy for whatever you want. mostly nobody. if you shrink the population by an order of magnitude (or even 2-3) all who remain will be happy
>>17060751>multiverse is made-up crapagreed .. i'm with penrose on this .. universe means everything .. that does not mean that there aren't layers of existence .. the whole thing is unified though imo
>>17055094It's over for science.
bump
>>17058589> AI is an eurekamachineNo it’s not> It can calculate the best medicineNo it can’t > but also the worst bioweaponsDitto> bioweapons is how humanity can be destroyed by AINo
>>17059390> rubbish. latent viruses can be developed atm, and have them trigger at more or less the same time with (seemingly) benign everyday use chemicals. like toothpaste/soap or whatever the point is culling enough so whatever remains chooses to be obedient anyway.Pure fantasy nonsense
>>17055094Science is done.
>>17056707The only reason democracy exists is because the rich need us plebs to work, when they don't need plebs to work what happens to all the plebs.It's not going to be T-1000s killing everyone it'll be thirdie levels of poverty with starvation being a leading cause of death.
