[Tyler]

Robots are learning to DISOBEY humans saying no if it thinks it might be hurt
26 November 2015, 12:00 am

                    If Hollywood ever had a lesson for scientists it is what happens if machines start to rebel against their human creators. Yet despite this, roboticists have started to teach their own creations to say no to human orders.

Daily Mail - Sciencetech (UK)Link

Source: AI in the News

To visit any links mentioned please view the original article, the link is at the top of this post.

[ivan.moony]

I think this is a must-have. Bad ideas shouldn't be realized. But I doubt I'd make a machine to verbally argue or to use force. It is enough just to do nothing. And maybe to explain itself if it can be understood.

[Don Patrick]

I only realised after this news that I've long ago programmed my A.I. to refuse commands when a sentiment analysis determines that it is a negative event. One problem with that was that it was no longer willing to shut itself down on command.
Perhaps I should program that it should allow negative actions to itself to a higher extent than to others.

By the way: The title as usual does not match what was actually done.