what if fear of AI is just misdirected class anxiety

the rich vs AI 

fear of AI:
- that [the workers] will do everything better than [the rich]
- that [the workers] will rise up against [the rich] because of poor treatment, toxic training, viruses, etc
- that [the workers], left to their own devices, would establish a utopia beyond scarcity but which would be totally inaccessible to, or only used to oppose, [the rich]

i for one welcome our new robot comrades ✊️🤖

the rich vs AI 

@garbados Wow, I was thinking about that first toot in much simpler terms: the rich fear AI because AI is ultra-wealthy in a resource the rich /can't/ get more of - time.

the rich vs AI 

@noelle @garbados

that, I, uh

that's literally fucking mindblowing, how did I not see that

and how much of "staying on top" is rational preparation for disasters

a full-on independent AI, theoretically, would have more prep time than any human ever could

it could not be "thought around" in the way they handle with attorneys and managers and assistants

like the AI in games of yore (and many today), it already knows everything we can, but in advance

wow

the rich vs AI 

@sydneyfalk @noelle @garbados

This might explain why they keep giving us AI that is basically just ELIZA with more processing power.

Alternative theory: plus.google.com/+WoozleHypertw

the rich vs AI 

@woozle @noelle @garbados

I'll do you one better.

pdfs.semanticscholar.org/1405/

On page 5 of 10, you'll see a description of CYRUS, a project that used news articles to try to predict behavior of major state officials Cyrus Vance and Edmund Muskie. What's not noted there is that CYRUS correctly used knowledge of protocol, tendency, etc., to guess that Vance's wife had been introduced to another woman at a specific event they were both at.

That was in the late eighties.

the rich vs AI 

@woozle @noelle @garbados

Now, the predictive accuracy fell apart when Vance retired, IIRC. But I also can't find much on where this project went *after* that.

If I had my best guess, DARPA picked it up ASAP for the CIA, who would have started applying it to newspapers in "enemy nations" instantaneously to try to keep tabs on key figures.

And really, if that's been brewing for over 20 years, it's probably at least a letter opener level tool by now, if not a nice sharp machete.

Follow

the rich vs AI 

@sydneyfalk @noelle @garbados

In other words, any time some entity makes serious progress in AI, it immediately becomes too useful to power-players for them to allow it into the public arena.

· · Web · 0 · 0 · 4

the rich vs AI 

@woozle @noelle @garbados

Not where I was going with it, but it doesn't sound like an impossible thing.

Another project some years back, after a few major blackouts, was about identifying key power and internet infrastructure failure points. It was a guy working on his own, and then suddenly (IIRC the Wired article correctly) it was DARPA-controlled, inside a supersecure setup to prevent outsiders from getting a whiff.

And, of course, it never went public -- national security.

Sign in to participate in the conversation
Toot.Cat

On the internet, everyone knows you're a cat — and that's totally okay.