top 39 comments

sorted by: hot top controversial new old
[–] 100 points 8 months ago (3 children)

"SWATGPT you killed the wrong person"

"You're absolutely right, that was my bad and I'll do better next time"

  • source
  • hideshow 3 child comments
  • [–] 27 points 8 months ago* (last edited 8 months ago)

    More like

    "You are absolutely right. However in an older version of the field document this person is labelled as a terroist."

    "Can you provide me that document?"

    "Sorry that document does not seem to exist anymore"

  • source
  • parent
  • [–] 55 points 8 months ago* (last edited 8 months ago) (2 children)

    We will build technical safeguards

    We all know that 1) they will rush it into use before that and 2) their safeguards don't work

  • source
  • hideshow 2 child comments
  • load more comments (1 reply)
    [–] 50 points 8 months ago (2 children)

    So Anthropic was requesting the DoW to include two (and only two) prohibited uses in their contract:

    • Autonomous killing with no human involvement
    • National, mass surveillance

    My understanding was… that’s all Anthropic was asking for, any other use was acceptable. That’s why the deal wasn’t going through.

    Now, Anthropic’s contract has ended, and the same day OpenAI is the DoW’s golden child?

    Something tells me OpenAI’s contract does not have those prohibited uses in it.

  • source
  • hideshow 2 child comments
  • [–] [S] 9 points 8 months ago (1 child)
  • [–] 12 points 8 months ago

    Amodei said in an interview that the DoW altered their contract to appear to compromise, so that it looked like they were allowing Anthropic to have those limits. But that legalese accompanying the updates rendered that text pointless. Basically, "We won't use Claude for domestic surveillance and full automated killing, unless we really want to." My guess is OpenAI signed the exact same contract and just pretended not to understand the distinction.

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

    Sam Altman: "I will gladly kill for money."

  • source
  • hideshow 1 child comment
  • [–] 49 points 8 months ago (2 children)

    And this right here is how Sam Altman survives the AI bubble bursting.

    Now they're a part of critical national infrastructure.

  • source
  • hideshow 2 child comments
  • [–] 2 points 8 months ago

    lol, no. their cash burn is still completely unsustainable. Altman himself will be fine when music stops, but the whole endgame with these tech grifters is bagging the US taxpayer (which'll happen when banks start going underwater, and have foreclosed assets they have conveintent buyers already lined up

  • source
  • parent
  • [–] 44 points 8 months ago (4 children)

    The DOW displayed a deep respect for safety

    lmao what

  • source
  • hideshow 4 child comments
  • [–] 15 points 8 months ago (3 children)

    Dude. It's SAFE War. The government promised?

  • source
  • parent
  • hideshow 3 child comments
  • [–] 34 points 8 months ago

    The world is a complicated, messy, and sometimes dangerous place.

    And you just made it worse. Congratulations.

  • source
  • [–] 21 points 8 months ago (2 children)

    Deep respect for safety? When I was in the army they made us do all sorts of wild unsafe shit. And we were worth millions to them via our training. How safe do you think someone who they haven't spent money on will be?

    Also the department of war jingoism is fucking annoying. Can't wait until adults are back in charge of the military.

  • source
  • hideshow 2 child comments
  • [–] 17 points 8 months ago

    The world is a dangerous place, so I'll be here to make sure it gets even more dangerous!

    Fuck you Sam, burn in hell you Nazi wannabe

  • source
  • [–] 11 points 8 months ago

    "to achieve the best possible outcome"

    For who? Even the people you're murdering?

    "Ensure our models behave as they should"

    We built a box that says "kill all brown people" and will act shocked that that would be it's solution to everything. It's so SMART!! It MUST just be the case that Muslims are violent and bad. Why else would our AI be so good willing to exterminate village after village?

  • source
  • [–] 10 points 8 months ago

    Some pieces of wood, a rope, and a sharp blade is the solution to all our problems.

  • source
  • [+] 9 points 8 months ago (3 children)
  • [–] 8 points 8 months ago

    War is peace, freedom is slavery

  • source
  • [–] 8 points 8 months ago

    So full of shit.

  • source
  • [–] 7 points 8 months ago

    Serve all humanity through war?

  • source
  • [+] 4 points 8 months ago (1 child)
  • [–] 4 points 8 months ago

    That insufferable scumfuck.

  • source
  • [–] 3 points 8 months ago
    [–] 2 points 8 months ago

    Lol they even used their made up fantasy name 🥾👅

  • source
  • [–] 2 points 8 months ago

    Oh no domestic mass surveillance? I don’t believe you…and if I did, you’re still happy to help mass surveillance everywhere else in the world?

    Man, I wish I knew how to organize people against these ghouls.

  • source
  • 9 comments in c/BoycottUnitedStates@europe.pub view that post
    [–] 19 points 8 months ago

    He's calling it the department of war.
    No need to read on, that says it all.

  • source
  • [+] 10 points 8 months ago* (1 child)
  • [–] 4 points 8 months ago* (last edited 8 months ago)

    The phrasing is actually non-committal. The promise is too big to mean anything or be actionable. They love to throw hifalutin' shit like that around because it costs them nothing. Never have I heard one of them specify what AI can really do, and how exactly it "serves humanity".

  • source
  • parent
  • [–] 5 points 8 months ago

    "We also will build technical safeguards to ensure our models behave as they should"

    You... you mean you didn't have technical safeguards before???

  • source
  • [–] 4 points 8 months ago

    Ol' Sammy saw an opportunity for some of those cash-packed government contracts and leapt at the chance, heedless of the fact his product will (will) one day slaughter non-combatants. After all, dead civilians can always be shovelled into the 'acceptable losses' column, can't they?

  • source
  • [–] 4 points 8 months ago

    Amazing how their "AI safety" efforts never mentioned how dangerous it is to let a private for-profit Psychopath leader/organization have control of a super-mind, or what would happen if said Psychopath gives that power to the biggest Fascist Psychopath sh*t-nation in world history..

    Boycott the scumbag, boycott US and all other tech-lords, or just rich/powerful Unts. Boycott all their favorite/recommended ideologies - they work only for them and create sh*t for everybody else !

  • source
  • [–] 3 points 8 months ago (2 children)
  • [–] 1 point 8 months ago (1 child)

    Instead of attacking a technology, perhaps attack the powerful Capitalist shitters that misuse that technology. Here we have an obvious Psychopath that are a direct cause of harm. A perfect target for your anger, yet you still chose to attack the tools they use. Come on man..

  • source
  • parent
  • hideshow 1 child comment