#eschatron

Asking for a friend.

The Office has the position that AI research and development should be tightly regulated. There’s about two dozen true “upper case” AI that we know of on the planet, so not that many, but each and every one represents a huge unknown variable.

I’ve touched on it before, and I’m not AbTech so this is my layman’s understanding, but “upper case” AI are nearly impossible to intentionally develop. All of the ones we know of had some sort of X factor involved in their development that brought them to true sentience. MISSI.exe had the psychic pressure of tens of thousands of teenagers, ALICE had a decade or so to digitize LSD molecules, CANTRIP was created by wizards, etc.

We have theories on what this means or how it works, but nothing solid. The Office has not attempted to intentionally create a sentient AI, but groups like AbSci U have and they’re only partially successful. That’s what makes them unknowns.

Oh Eschatron has LEP and all that. I don’t want him destroyed.

Keep in mind that he was also built and programmed to actively seek the end of the world and/or universe. At all times he is the million monkeys banging on typewriters, but instead of writing Shakespeare he’s attempting to snuff out all life on the planet. At a certain point the Office is willing to endorse shutting him down at least until we or AbSci can figure out how to reprogram him - if they were willing to do so, which they aren’t.

Important question: when you say “living in it” do you mean a demonic or otherwise extranormal entity, or do you think it’s some kind of AI?

Contrary to what you might think, the former is much more likely. True, capital-AI artificial intelligence is extremely rare. There’s been about a dozen that the Office can say with confidence pass a sentience-level Turing test. I think three of those are in our custody; ALICE, MISSI.exe, and CHARLEMANGE. There’s ETCetRA’s CANTRIP, AbSci U’s Eschatron, RASPUTIN….I’m probably missing a few, but the point is there’s few enough that I could just name them off.

Oh no, the geniuses at AbSci University don't have any limiters on Eschatron. They said it doesn't preserve the scientific merit of the "experiment". If they fail, it's up to us and the other allied orgs. Granted they're monitoring Eschatron 24/7 so it's incredibly unlikely it ever succeeds, but you can understand why we want them to shut it down.

Not remotely close, which is good. So far ASU has done a good job of keeping him contained and I will admit that his data helped with the ███ ███████ ████████. I wouldn't worry any more about him than you would about a meteor.