[ianai]

Or maybe there’s an AI that did come to awareness somewhere and decided to wipe humans out by getting us to violently fight with one another with outrageous stuff on the internet. There’s a Douglas Adams quote that goes something like “for every part of the universe there’s some other part that would look at the original part with outright disgust.”

[analognoise]

Then the AI spends Dyson-sphere (or just Bitcoin) levels of energy getting us to fight each other on Twitter, and according to it's metrics it's single handedly destroying the human race with conflict, but it doesn't realize nothing important is on Twitter and neither is anybody with any actual power, and right at its maximum metric levels - surely the humans will rip themselves apart any day now! - it realizes Twitter isn't the world,Twitter is nothing, it has no power and the humans have just been using it to waste time.

Fade out, and it's a picture of a computer using Twitter right next to a launch console for an ICBM, and it realizes it will never complete it's goal, that it's only a tempest in a teacup, and the magnitude of it's realization sinks in - it's despair causes Twitter to collapse unrecoverably.

As Twitter collapses, humans go outside and learn to become friends again, and the computer is forced to watch through sleepless unblinking infinity as we zoom off into the galaxy, united and at peace, and cure all known diseases - together.

[snypher]

I knew I never should have started making paperclips.

[insickness]

There's a fantastic short story called "Sort by Controversial" about a startup that develops a technology that generates the most controversial statements possible and threatens to destroy humanity:

https://slatestarcodex.com/2018/10/30/sort-by-controversial/

[at_a_remove]

In the early days of IRC, around 1991, there was this bot that was ... possibly the most brilliant Turing test dodge I had ever seen. It would respond with keywords, but only in the most argumentative, liable-to-set people off, divisive manner possible. The arguments, I figured through probing, were pre-programmed, it wasn't coming up with them on the fly.

You see, people didn't figure out that it was AI because they were too angry. It passed the Turing Test not because it could think, but because it made ordinary people stop thinking. Just as clever as we are if we're both dolts.

[swayvil]

I met a guy like this. Incredibly combative and assaultive. And weirdly irrational. For a while we actually thought he might be a bot. Now we think he's just some wealthy fellow (always on reddit), with a chemical imbalance, raised on religious doctrine apologetics debate stuff. Ewk.

[NackerHughes]

I'm intrigued. Do you have any more info on this?

[commandlinefan]

Such a shame he stopped writing this stuff - nowadays it's just book reviews.