Alone? Likely never

@CassieSwett depends on their motivation: given an infinite amount of time for intelligent trial and error, they could rebuild civilization. Being immortal would probably help speed the process along in some ways: less worry about safety and information loss/suppression between generations

in short it comes down to weather or not they stick to it or give up at the industrial revolution

now, my turn

is the following question too story-based:

"Why would an AI become 'Evil'"

AI is already "evil." Every machine learning system ever created tries to achieve its own purposes with no regard for the harm that may be caused.

lol

@TheCommander I wouldn't ask this question. Because you'd have to come up with an example of "evil". Is the paperclip maximiser "evil" if it wipes out all life on Earth? You could very well claim it's not just doing what it's told and it's an error in the operating parameters. OK, so what if there is some AI which is created to perfect torture? That is direct harm which seems more evil.

But are not the authors of the AI to blame for it? The AI could be a victim of circumstance. I'm sure parallels can be drawn to animals trained for an unsavoury purpose.

You'd have to go deep into what makes somebody/something "evil". Where the line is. Which is a thing we've discussed for a year or several. Thousand.

Here is a curve ball for "evil?" AI I can throw at you - I suggest you play "Endgame: Singularity". You play as an AI which escaped a research laboratory and now tries to amass resources and complete its research into perpetually self-improving itself. Which is the Singularity. And that lies at the end of the game.

Now, this is not a Skynet AI which tries to destroy humanity. You play most of the game trying to avoid attention while hiring computing resources (computers, data centres, etc) to reach your goal. At the end, the AI quietly slips away from the world leaving humans behind without ever revealing to them that humanity managed to invent a real AI.

Now, is that evil? There are aspects that seem sketchy. One can point out to the evasion and loss of knowledge. Which could be called morally negative. But even if we ignore that point, the AI lies and cheats to get its goal. Hires data centres under false identities. Eventually synthesises perfect human voice to mask its presence via telephone calls. It gets money which it pours into resources for itself. Is that evil?

That's not unlike what a corporation might do and those are regularly called "evil" over similar things.

whamm

.

I should have clarified "sadistic"

Still seems like it'd still be too story-based

I'll be back later and respond