"...how to ensure people can interrupt an AI system whenever needed, and how to make such systems robust to cyberattacks." Good luck getting both these right. Who does the AI allow to interrupt it? Who gets locked out? How does the AI tell the difference? I doubt AIs are going to get really good at this until they are so good they can make up their own minds which people are allowed to "interrupt" them.

If your objection to homeopathy is that it is ridiculous because it couldn't possibly work, you are simply repeating the silly style of objections of everyone who has every misunderstood the complexities of the universe. The world is clearly flat, obviously. The question isn't whether you can believe that it works. The question is whether it works. Nothing else matters, except as a matter of interest. Condemn thee not what thou dost not understand, lest the future prove thee foolish. Accept the possibility of those things that appear to be real, but which defy explanation. Otherwise, be a slave to your own preconceptions.

Begging the question dear.

Breitbart.com? What has the world come to? Or slashdot for that matter? I'd need to deslime my eyeballs if I followed that link. I'd need to clear my browser history so I could forget. Yuck.