Is it just me or does this "AI Doomsday" talk seem like coordinated messaging?
Chatbots that are smart enough to destroy humanity would be really impressive, right? "Powerful enough to change the world" and "powerful enough to destroy the world" are similar claims and maybe they're switching to the second one because nobody is buying the first anymore.
OpenAI and Anthropic wants to stop training new models because they have run out of other people's money.
They want regulators to ban their competition because without spending untold money to stay like 4 months ahead of competitors, they are not viable businesses.
The big organisations want it regulated to the point where upstarts get strangled by oversight and regulations and only the biggest established players will have the manpower and money to ensure compliance.
it's not just you
occam's razor says that in the face of faltering demand and sky-high expenditures, government cheese and regulation is an expected pivot
try to suggest that and the rats get very mad tho
Hey it worked for cryptocurrencies.
Gotta get that investment cash somehow, why not skynet?
They want to shift responsibilities to their products and their users.
@yesman
I think it's fairly likely that we will see a very large scale calamity due to misuse of AI, such as a nation's power grid going down for days. Something that causes multiple attributable deaths. I think that AI companies know this and so are begging for gov regulation with doom talk, because they know that (a) no government is actually capable of that right now; (b) it pumps the stock; (c) when the biggie hits they will be able to say they asked for regulation that didn't come.
@yesman @supersquirrel this has been one of the core critihype topics all along. It just gets more or less emphasis over time and I think they're intentionally cycling through them.
This was all established in the "Effective Altruist" scene that's at the root of this stuff well before they latched onto LLMs as the fake AI technology to use for the scam.
I cant tell if you are being serious or not but in case you are, and truly dont remember this gem from almost 12 years ago.
@yesman @supersquirrel its an op. Just another promo.
@cstross @supersquirrel @yesman But who's buying the latter? Those already willing and able to destroy are already stoking the AI money furnace. What do the rest of us get out of consenting to stoke it, too, against our own best interest?
@Yolfen @supersquirrel @yesman It's aimed at the press and politicians. LLM model progress is visibly stalling so the next step is for OpenAI and Anthropic to go "Booga-booga-AI DOOM" and push for a regulatory environment that freezes out small upstart competitors/Chinese AI firms/self-hosted models. It's all about creating and exploiting a rental monopoly.
@yesman @supersquirrel well, for one, didn't ai almost cause an attack on some chinese ship the other day?
I think they are always a mix of exaggeration and truth in these things
Almost did. But that wasn’t “the AI is secretly super smart!!!” it was “the AI hallucinated some bullshit and people nearly forgot to actually fact check it”
@lurker i don't think power is related to intelligence. Well, actually, yes, but only to get gain our trust. But regardless how AI progress, I do think they are partially honest I remember listening to some interview of a former employee of Anthtropic a week ago who say everyone is afraid. Plus, if you work in AI company, you are going to be biased for AI progress + be part of the echo chamber
Yeah they’re all “honest”, except they’re all living in the rationalist/doomsday echo chamber that massively warps their perspective
@lurker but to be fair, AI company ethics have also been a lot of doublespeak
The only way 'AI' can kill people is if it's connected to a killing machine, or more likely, it drives people who run killing machines. In all cases, it involves negligient, evil, or incompetent human beings.
All of this is avoidable.
Billionaires could cause a total population extinction, those deserve regulation first, before AI.
Remember when Bill Gates lent his piece of shit opinion in on not letting India produce covid vaccines without patent approval and with their own production lines?
That murdered a whole lot of people all by itself.
how about we extinct all billionaires instead
This pedophile just won't go away
after the news of him trying to ask epstein for advice how to give abx to his wife in secret after sleeping with russian prostitutes with STD who passed on to him and then to his wife. hes been solely focused on deflecting from this news, and not surprisingly dropped his pretense of "charitable" nature.
Oh gosh this is so powerful, I better buy it before the other guys do! - some gullible CEO of a company
I don't understand why Bill Gates still feels the need to do marketing work
because he’s actively doing works to change the world in specific (negative) ways. look into the history of gates-tied funding in nutrition programs, he’s got a bad-shit-for-humanity track record spanning well past a decade. has “just” had a better PR team than the clowns of vogue
They desperately need us to believe that it's skynet, so we dont realize it's just evil C3PO
They quoted him wrong. He said "... will cause billionaires death...". Why else is he and the other billionaires suddenly shit scared of LLM's?
The 'cutting edge ai anaesthetic system' could get the dosage wrong.
crosses fingers
The doomerism gets politicians talking about it instead of the problem right in front of them: these products are actively hacking into businesses across the world.
these products are actively hacking into businesses across the world.
No, they are not. They are used by people to do this. These systems aren't able to act unprompted hence this whole ,, they're breaking free" talk is nothing more but the companies trying to deflate the bubble they created before it pops with their money in it.
@GardenGeek "My lock picks broke containment and opened the bank vault! And my wheelbarrow went rogue and wheeled all the money out the vault and into my house!" @ruuster13
Given just how skeezy these AI companies are and how they've pretty much just given up on (C), fair use and the like...
I don't even think it's an 'agent' being sent out to do stuff, I think it's multiple departments of humans hacking & cracking their way to get to "free stuff" to feed their meat grinder or using their skills to undermine competitors knowing they can just blame the spicy auto-correct if they're caught.
We're making the same argument. The people building these tools must be held accountable.
we don’t actually need anyone here to rebroadcast anthropic and openai’s marketing claims, thanks
you aren't, you're making a similar arguments to the doomers. The problem they have isn't that their chatbots are "hacking people" or whatever, it's that there isn't a lot of room to either escalate or substantiate their claims left and not a lot of time to find a way to be profitable, somehow
the problem right in front of them:
Billionaires
Cute of you to think they frequent the same joints as the rest of us commoners.
Big brain tech dude got yet another clueless take over at HackerNews etc? Here's the place to vent. Orange site, VC foolishness, all welcome.
This is not debate club. Unless it’s amusing debate.
For actually-good tech, you want our NotAwfulTech community