The three 'Laws' of robotics would be extremely hard to write, as the supervisory system it was written to be!
And in the entire collection of his work, the robots actually wrote a 'zeroth' law that single people could be sacrificed to save humanity in general....
AI can be a good servant in limited use. I need to repair a gas fired oven (about 12 by 20 feet); because the operator decided to speed up the lighting process, by setting the AI controller to all zeros! The explosion was small, and we have an impressive fire department. So not all AI's are bad. That said, DARPA concerns, as they have no feel for risks related to their projects! They have tried for several decades to locate a shooter by triangulation of echoes. I have this problem figured out, they are just doing it wrong! My dilemma is, do I do trust them with this ability? Will they use it against innocent cent people? If I told them their mistake, they could have an accurate, portable system in less than a year. They would pay me a lot for what I have; but I don't trust them to use it morally!
AI, when used in limited Small control applications, it can beneficial! Big systems can become dangerous when they grab too much power. I'm with you, post what you fine he and I'll help!
Hi Toby
Thanks for the comment, nice hearing from you. Gosh that explosion sounds terrible and could have ended badly but I'm glad you got it under control.
This is the main issue and really comes down to morality as you say : "but I don't trust them to use it morally!" I think that we are always going to have those that will use things like this for nefarious purposes.
I listened to a Ted Talk on this a few nights ago - an old one, posted 5 years ago already and the main thing that scared me was when he told the audience that AI evolves and updates itself: it writes it's own code for this which then throws the developers out of the loop. How can this not be seen as something frightening? Where are the checks and balances? I see it spinning out of control in the future, but I'm just a really cynical old bat. When I first heard about Bitcoin, I thought it was a bad idea too, but that's because I didn't know anything about blockchain tech.
Will be interesting to see how this goes in the future.
As for DARPA, don't trust those mofos at all!
Yes, unsupervised systems are a danger, and the same system that is hacked is even worse. Some dark hacker could deliberately redirect an AI to be deadly to organic life!
It's like fire, it makes a good servant....
They are still rebuilding the be damage, so I haven't had a chance to restart the AI yet.
DARPA would pay a lot of money for what I know, but I lost faith in their morals long ago! Sad too, because it is an elegant solution; to a difficult problem....
Maybe I will make one for my own use.
👍🎯💙😁