I saw someone give an interview the other day – saying that AI will eventually control us – people in some way
All makes perfect sense
: It will put other humans to work to serve that higher interest
And at some point a human might give it a goal – I suppose – to serve itself – rather than any obvious human interest – or to achieve some abstract goal – that may be shared by humans – but which is not in any real sense designed to further the goals of a particular human…
At which point – all humans may become subservient to – or irrelevant – or a nuisance to the AI goal
I guess it really kicks in once AIs are able to control and co-ordinate the actions of loads of robots – which have a physical presence in the world…
Then it all becomes very matrix…