5 Filters

AI Grok2 programmed to kill billions of non Jews to save one Jew losing 50 cents

Do we want Starmer to let this happen!

Cheers

2 Likes

This implies the worth of black lives (or any other non-Jewish lives, as shown by the other questions and answers) is zero.

And…a robot so programmed with these values would, for example, happily shoot a black man for whom it reckoned there was a 10% chance he would steal a Jewish man’s $5, or takeaway food.
It might not due to overrides implemented - but that’s what it would want to do.
Even stopping short of certain visible riot-triggering violence, still about the worst fascist you could imagine.
Clearly the AI is like a policeman’s baton; inanimate, but dangerous in the wrong hands - and hard to imagine the right hands.

1 Like

I may have missed the main point…they can paper over the cracks such as this (they will call it an aberraiton, a bug) but then underlying values will remain. These answers - even once removed by an override of some kind - show what chance you have of achieving a fair results against the set of values, in any scenario. You might not be shot but you and you’re problem are still assigned a value of useless (if you behave) in the value system.

I think the ‘hand’ signal in the middle of this image is fairly representative of my feelings about Musk, Nuttyyahoo, and AI generally.