[pandapower2]

>If I had to guess the next touchscreen everywhere will be misuse of AI, not in a malicious way, but in a really stupid one.

Reminds me of a talk someone gave. I forget who. Elon Musk perhaps. A quick google didn't turn it up.

Anyhow, they laid out a hypothetical scenario where humans were driven to extinction by a run away AI. Rather than being some sort of Skynet scenario where the machines are explicitly trying to kill all humans, it is an AI whose job was to maximize strawberry production. The AI converted the earth's entire surface into strawberry fields and inadvertently wiped out everything else in the process.

Arguably a more plausible scenario than Skynet.

[iofiiiiiiiii]

The Revelation Space book series by Alastair Reynolds [spoilers ahead] ends with such a paperclip maximizer scenario. Someone accidentally turns on advanced sentient self-replicating terraforming robots without programming their target area. They promptly begin to terraform the galaxy and don't take kindly to interference.

This is a tiny shade of background in the story, not the main story, but read the books if you like this kind of tale.

[freerobby]

That talk is a spin on Nick Bostrom's Paperclip Maximizer. See: https://wiki.lesswrong.com/wiki/Paperclip_maximizer

[yyx]

https://en.wikipedia.org/wiki/Instrumental_convergence#Paper...