I have been very confused so far this year, that every other week we hear about some company/country being attacked by an AI, but then everyone going “oh well, nothing anyone can do anything about. What a pity”.
This is 100% the case. If an engineering firm fucks up and a bridge of building collapses or a CPU lacks working floating point arithmetic, they’re going to pay out the nose for it.
They’re running around pretending they can’t control any of this, but without giving a language model access to a shell or something, what the fuck can it do?
We have to deliberately grant these abilities. Otherwise, they don’t exist.
HUMANS are accountable for 100% of this.
Start slapping cuffs and it stops. They want regulation - regulate them into 6’ cells
I have been very confused so far this year, that every other week we hear about some company/country being attacked by an AI, but then everyone going “oh well, nothing anyone can do anything about. What a pity”.
This is 100% the case. If an engineering firm fucks up and a bridge of building collapses or a CPU lacks working floating point arithmetic, they’re going to pay out the nose for it.
They’re running around pretending they can’t control any of this, but without giving a language model access to a shell or something, what the fuck can it do?
We have to deliberately grant these abilities. Otherwise, they don’t exist.
That’s not quite true. As with the huggingface attack, the ability can be negligently granted as well.
The responsibility doesn’t change.
I took an oath. In tech. I take it as fucking seriously as any doctor.
Oh yes, doing something negligently instead of actively absolutely does not absolve you of any responsibility.