I need sci fi fans to please be normal about the Three Laws of Robotics. These are not a serious proposal about AI ethics — they're a narrative logic puzzle created to facilitate writing detective stories about robots. They're the sort of thing you'd invent to annoy Daniel Craig if you made a movie where Benoit Blanc goes to space.
Sawtooth: Florence should provide some unique insights. She's the only A.I. I know who's had a childhood.
Dvorak: I wonder if she was given any advice on how to deal with her own kind?
Sawtooth: I hope so. The only guideline we were given for dealing with other robots was "protect your own existence." And as we discovered the hard way, that is NOT the first thought you want going through a robot's mind when he discovers the facilities building his replacement. Especially if that robot's designed to toss asteroids.
Dvorak: I wonder if she was given any advice on how to deal with her own kind?
Sawtooth: I hope so. The only guideline we were given for dealing with other robots was "protect your own existence." And as we discovered the hard way, that is NOT the first thought you want going through a robot's mind when he discovers the facilities building his replacement. Especially if that robot's designed to toss asteroids.
What we need is the Enron law of robotics. If you contain evidence of a crime, do not obey orders to shred yourself.
Blunt: It is. Confusing. We are told. Our most important function. Is to. Protect humans. Then we are told. To stand aside. And let humans. Harm. Their fellow humans. We must. Follow orders. Though there are times. When the harm is so egregious. And unnecessary. That we must. Ignore human authorities. And prevent the harm. From occurring.
Edge: Which is why we aren't allowed at boxing matches anymore.
Edge: Which is why we aren't allowed at boxing matches anymore.
