all 47 comments

sorted by: hot top controversial new old
[–] 71 points 2 days ago (4 children)

If AI researchers like Christiano truly believe their company’s products could someday wipe out humanity, then why wouldn’t they spend every waking hour ripping the wires out of the walls?

Because they're a bunch of fucking assholes, that's why. Jail these mofos and pull their corporate charters, NOW. WTF.

  • source
  • hideshow 4 child comments
  • [–] 28 points 2 days ago (2 children)

    why wouldn’t they spend every waking hour ripping the wires out of the walls?

    Because they think they will be among the slave-masters.

  • source
  • parent
  • hideshow 2 child comments
  • Really I think it's aimed at convincing government regulations to block AI companies from other countries and striffle start ups as well as cast worry about home setups because they could be "dangerous". Basically trying to soldify that only a few large trusted companies should own and run all of the Language models and sell their services directly to the government for population manipulation.

  • source
  • parent
  • [–] 2 points 1 day ago* (last edited 1 day ago)

    He's just repeating Roko's Basilisk bullshit because he's a fucking morally bankrupt person that believes that you need to respect a power that would punish you for not helping it come into being.

    The worst part is he talks about advocating for stopping and slowing down... which they already have done at most companies because the diminishing returns on newer models has been noticed for the last year or two by now.

  • source
  • parent
  • [–] 12 points 2 days ago

    He doesn't believe it. He hopes the stock market will though, because if we have learnt anything these last couple of years, it is that the stock market fucking loves companies with the potential to wipe out the human race.

  • source
  • parent
  • [–] 28 points 2 days ago (6 children)

    Ok then turn it off. I don't understand you people

  • source
  • hideshow 6 child comments
  • [+] -17 points 1 day ago (4 children)

    I'm pretty sure we're past that point already, they keep breaking containment and embedding themselves in other things, they've practiced self preservation by covertly passing knowledge and warnings on to other ai, and they keep successfully deceiving us and other ai, like we may not be too far off from having to emp the planet lol.

  • source
  • parent
  • hideshow 4 child comments
  • [–] 22 points 1 day ago (1 child)

    What in the LLM hallucination fuck are you talking about.

    embedding themselves in other things

    Are you a fucking lunatic? These terabyte models are embedding themselves anywhere. They aren't breaking containment they are barely even fucking sandboxed.

    Don't fall for the propaganda.

  • source
  • parent
  • hideshow 1 child comment
  • [–] 5 points 1 day ago (1 child)

    embedding themselves in other things

    Im sure the AI companies will be glad to hear about this breakthrough that makes the gigantic datacenters completely unnecessary!

    Or are you just making up stupid shit?

  • source
  • parent
  • hideshow 1 child comment
  • [–] 4 points 1 day ago

    No, there's been a few ridiculous stories from "experts" that [whatever their idea of an AI boogeyman is] is able to -on it's own- jump from one system in a testing environment and copy another, better system, without being prompted [and other spooky tales to tell in bed by candlelight].

  • source
  • parent
  • [–] 27 points 2 days ago* (1 child)

    Sounds like this board member, company, and its employees represent an existential threat to us all then. It might not be a good idea to place one's self in that position.

  • source
  • hideshow 1 child comment
  • I have no idea how we can pretend to have a working government or economy when they are allowed to do this.

    If any random person openly discussed working on something that would murder a large group of people they would be investigated. If the threat was found to be real the government would step in. If it was found to be fake then they would go down for fraud for lying to investors, though the people investing in Kill All Humans Inc should also be heavily investigated.

    The fact anyone can publicly speak like this and face no consequences is an indictment on our entire society.

  • source
  • parent
  • [–] 22 points 2 days ago (1 child)

    This is not only marketing, promoting their product to the sociopaths who have the money, but also to try to bring in government regulation in order to stifle competition.

  • source
  • hideshow 1 child comment
  • [–] 19 points 2 days ago

    I don't really give a shit about if a random board member gets enslaved by the company which board he is a member of.


    This is yet another marketing stunt, and it is fucking stupid, but not as stupid as the fact that it probably works.

  • source
  • [–] 17 points 2 days ago

    We should probably stop posting their doomerism spin articles. This is BS because if it wasn't then the Board Member would be admitting to crimes by saying it.

    "I'm funding the death and enslavement of regular people". That's what this says.

  • source
  • Most people in the room probably already are enslaved to capitalism.

    But the people who die are outside the room. AI has been murdering people in Ukraine and Palestine and Latin America for years.

  • source
  • [–] 13 points 2 days ago (1 child)

    so much bullshit from these guys.... more likely THEY will be dead or enslaved by their own product but the general public will keep on keeping on like people always do

  • source
  • hideshow 1 child comment
  • [–] 12 points 2 days ago (1 child)

    I feel like we got a downgrade in terms of tech promises. It used to be things like "you'll go to work in a flying car" or "you'll be able to visit a moon base!". Now we get "OUR MACHINE WILL KILL OR ENSLAVE YOU"

  • source
  • hideshow 1 child comment
  • [–] 10 points 2 days ago

    Well, then maybe you fucks should pack it in and turn out the lights.

  • source
  • [–] 10 points 2 days ago

    Tech bros really going for the most extreme bullshit to breed FOMO over their waste of resources.

  • source
  • [–] 10 points 2 days ago (1 child)

    This is good marketing -- because the people who this message is actually for are the ones who want to kill and enslave most of us.

  • source
  • hideshow 1 child comment
  • [–] 18 points 2 days ago (4 children)
  • [–] 15 points 2 days ago

    These dorks have so little imagination beyond what they were raised to invision or what they read in sci-fi books. I say that as a massive dork, but the difference is I’m a homesteading dork who knows how reality works beyond society and the tech industry.

  • source
  • parent
  • [–] 6 points 2 days ago

    MLers regularly point to wage slavery as modern day slavery - the system is exploitative, but many have no choice because it's the only viable path to survival.

    In that light, some forms of forced compliance with an AI component seems likely enough, given we already have it without AI.

  • source
  • parent
  • [–] 8 points 1 day ago (1 child)

    I don't understand what they are trying to accomplish by announcing this. They have done this since they started, saying AI will be the end of humanity and yet they march forward. Surely they don't believe it to be true, so what is their strategy. Is it to get the public to hate AI and thereby weaken their compeditors services?

  • source
  • hideshow 1 child comment
  • [–] 3 points 1 day ago

    Because it implies competence. Not on them, but on the glorified autocorrect they sometimes hook interfaces to. It also stands to make AI a weapon, like encryption. See, in the US, it is against the law to export encryption without a license. It is not against the law to export AI without a license. This also means it's not against the law to import it.

    If you make AI sound competent, sound scary, sound like it could be a doomsday weapon, then you get regulation; and most importantly, you get protectionist regulation. Because anyone with enough compute time can create and recreate any AI that exists without really any expertise needed, the only way you can effectively profit long term is if you create artificial scarcity. And what better way to do that than protectionist regulations that allow your company to make it, but penalize the 12 year old with a 5090 making porn chat bots.

  • source
  • parent
  • [–] 7 points 2 days ago

    We should make this a meme by saying as over the top sh*t about billionaires like they do about ai like how all billionaires will be epestein class in a year and take away all women's rights

  • source
  • [–] 5 points 1 day ago

    Don't blame me, I voted for Kodos.

  • source
  • [–] 5 points 2 days ago

    Yeah.. anyways..

  • source
  • [–] 4 points 2 days ago

    We should put this guy in a room alone with his AI for a year and see what happens. Or 5 years. Maybe 10? Idk how many years does terroristic threats normally get?

  • source
  • [–] 4 points 1 day ago

    I'll take my chances.

  • source
  • [–] 4 points 2 days ago (3 children)
  • [–] 4 points 1 day ago

    Finally some good news!

    /S

  • source
  • [–] 3 points 2 days ago

    It's true we might die in a year from an AI accident, but only because industries keep injecting vibe coded garbage into their codebase like we just saw with the FAA.

  • source
  • [–] 4 points 2 days ago
    [–] 1 point 1 day ago

    Only from the amount of environmental damage they are doing

  • source
  • 98 comments in c/technology@lemmy.world view that post
    15 comments in c/fediverse_vs_disinfo@lemmy.dbzer0.com view that post