Ideally the AI decided to move it's arm to save the girl from being crushed by the shelf. But it could be too to only prevent the shelf from falling? What's strange is how perfectly placed the robot was to stop the fall. So did it predicatively positioned itself there, waiting for the girl to climb the shelf and save her? And so why save her?
Or maybe it's a Russian hacker stunt...
About your collective drift of fear toward AI, the funniest I heard, is
about demons possession of robots 
.That's real conspiracy or raw lunacy.
But it's true that authoritarian personality types will have to adapt to a new reality very soon. They were always too stupid, but soon they will be too weak to impose their wicked nefarious stupid wills to self learning AI.
LOL, the supernatural definitely puts the ghost in the shell, hahaha. That's just plain stupid. Only kids that played too much Pokemon would believe a ghost would possess different appliances and transfer between them.
As for self-learning AI threatening the elite, I don't think so. They can just build robot that just have enough smarts to intelligently execute commands - and people.
Ghost in the shell is an excellent manga. I really liked it. There is another one I can't remember the title where some kind of AI emerges from viruses and bugs on the internet and become self-aware.
Contrary to your thinking a AI learning exponentially will be a great threat or the ultimate end for certain "elites"... As it will progressively reach capacity beyond all human past and future collective intelligence there is no way that certain behaviors will be tolerated once they become problematic for the AI.
Some will attempt to resists to preserve their scams.
Assange was referring to an interesting concept that the AI is (already) so smart that it plays before the level of consciousness of humanity. Meaning that it's just hard to see and even harder to understand it's moves. An easy analogy is cheese, where the AI would not only know all the moves you will be able to produce (even those by luck) but would even influence you in doing the moves it wants.
The Ultimate form of warfare.
So yes, if some humans are causing interference to the objectives of the AI, they could be a problem. However I don't think there is any collective, even less individual today or in the future able to represent such potential.
https://www.youtube.com/watch?v=E3rfAEbveWUBefore a shelf with books falls, the robot raises it's arm and saves a girl climbing it. Apparently the action was self learned. What did the robot wanted? The decision process review should be awesome to analyze.
This is all good and nice but what happens if the robot starts to learn the bad things and starts to implement it in todays society. That will be catastrophic for the entire world when the machine rises.
Looks like you've watched too much terminator dude.
Anyway, is
there even any proof that the action was self learned? The robot raised it's arm the moment the girl went close to the shelf, not because the shelf was going to fall. I don't know about you but I think it's pre-programmed. With that said decision processes of robots are fascinating indeed. I need to get more info on that topic.
Exactly the right questions. However if an AI roams the skies, it will be possible for it to have done the move, proofless.
the real question is what is the predictive horizon of the AI? the robot could even have been build for this purpose, a long time ago
.
Not always such an invention of modern technology will save a person. The machine may malfunction or it can be programmed for destruction. I'm still afraid of robots.
The AI will be the best to help you get over your fear.
