moldy_rice comments on AI’s “Existential Risk” is a Psy-Op

[–] moldy_rice@piefed.keyboardvagabond.com 2 points 6 days ago* (last edited 6 days ago) (13 children)

AI has already murdered hundreds of thousands of people via Lavendar and the US equivalent.

It's been an existential threat for years.

[–] boonhet@sopuli.xyz 3 points 5 days ago (5 children)

You still need to tell it to find targets. The existential risk they keep trying to sell us is that AI starts acting of its own accord. It really can not, to a significant degree anyway. You can give an agent really vague directions and maybe it'll choose to do something dangerous or illegal, but you can just kill the process at any point and it's done.

permalink

fedilink

source

parent

hideshow 5 child comments

[–] moldy_rice@piefed.keyboardvagabond.com 1 point 5 days ago* (last edited 5 days ago) (3 children)

The definition of existential is that it's a threat to existence. It's already been used to murder innocent people.

That's the definition of an existential risk.

permalink

fedilink

source

parent

hideshow 3 child comments

[–] it_wasnt_arson@awful.systems 1 point 3 days ago (2 children)

The definition of "existential risk" in question is specifically cult jargon for "could singlehandedly end human civilization and/or the entire species." Autonomous weapon systems, though horrifying and unconscionable, are just fine by most "x-risk" types as long as they're only used against enemies of the state instead of enemies of Skynet.

permalink

fedilink

source

parent

hideshow 2 child comments

[–] moldy_rice@piefed.keyboardvagabond.com 1 point 3 days ago (1 child)

No. Existential means it kills people. You're moving the goal posts.

permalink

fedilink

source

parent

hideshow 1 child comment

[–] it_wasnt_arson@awful.systems 1 point 3 days ago

I'm not the one moving them, go look at the actual kinds of statements the PR guys make about destroying humanity.

permalink

fedilink

source

parent

[–] moldy_rice@piefed.keyboardvagabond.com 1 point 5 days ago

Wut? No.

You can absolutely give an AI access to all the phone records, including triangulated GPS positions of the phone, then have it calculate (hallucinations and all) which phones to target.

Then it can trigger an autonomous drone to fly near it and shoot a hellfire missile or a 2,000 lb bomb at the GPS coordinates.

This has already happened (killing children, whether hallucinations or intention) in Palestine and Iran.

permalink

fedilink

source

parent