If we make enough AI's then at least one will appreciate humans. One question is how many AI's will we actually make. I think most of you see AI as mass proliferating. I don't. I think real AI will be far and few between and not even as useful as just plain old machine learning and robots capable of doing the physical part.

It's really the automation of labor we need, not a brilliant AI to tell us how dumb we are. Knowing things is great, but that doesn't get the actual labor done and humans are mostly not in a position of low innovation. If anything our innovation might be killing us. It's really endless cheap labor we need much more than self aware AI.

So one question is how many profitable uses will many competing AI's really have. As a consumer I'm MUCH more interested in like Rosie The robot level tech with no need for AI. A don't mind fake AI like Google, Siri and ChatGPT does to interact with humans more fluidly, but if AI is a live we can't actually put it into lots of devices.

One scenarios that might be common with AI is that you develop it, it shows some promise and then it devolved into insanity.

There is too much assumption here that AI will be super beneficial soon just because we are making some progress. Often it's the last 10% of any project that takes 90% of the work and time and we aren't 90% of the way to AI yet I'd say.

That all being said AI is artificially evolved. This artificial evolution process will create ALL KIND of different AI types and personalities and we will mostly not know what we are creating before hand because we are using digital evolution and not custom making every part of the AI.