>It could destroy society because what if it becomes self-aware and realises it doesn't need to tolerate humanity's shit!?
Why is this (or something similar) attitude the main one when it comes to whether or not AI is a good idea?

The reason why AI is going to destroy society isn't because of some SciFi plot from the Matrix, it's because AI will automate middle class careers, society doesn't have a response to that and seems extremely unlikely to come up with one in time. It's already obvious that most manual labour jobs that still remain will get replaced with mechanical robots, and people act like anyone who loses their job to a robot are at fault for "not trying harder at school". Well what is our main industries now in the west? Finance and Banking, Marketing, Investment, Management.... these are all industries that AI will revolutionise and put people who work in those sectors out of work, just like the people with lower paid jobs.

So if working and middle class people are both put out of work over computerisation of the industry, what are we doing to make sure that society can actually tolerate such a reality? Because I don't think "well if you didn't want to be poor AF, you should have been smarter than a supercomputer" is going to cut it.