3 ms·

I do hope we get there. In the short stories it was made clear that robots couldn't lie, and that they could prove it was impossible for the robots to circumven

I do hope we get there. In the short stories it was made clear that robots couldn't lie, and that they could prove it was impossible for the robots to circumvent the three laws (although they are on occasion incentive on how they interpret the word "harm" specifically). If an LLM couldn't lie and could be provable shown to be unable to do so would be quite powerful.

jprete 2y ago
The short stories ended with the robots firmly, and invisibly, in control. "You're not allowed to let humans be harmed by your inaction" inherently requires the robots to take over in whatever way causes the least harm.