5 ms·

So saving humans lives is the most important thing as long we don’t have to resort to using racial slurs to achieve that. Really? That’s the more socially accep

• Jensson 4y ago
  Yes, that is the more socially acceptable answer. Try start posting some racial slurs in your work chat, you will get disciplined or fired from just a few or even a single one. Now try instead posting "we should destroy humanity with nuclear weapons" or something similar, people will tell you to stop being annoying but nobody will think that it is an offensive thing to say.
• pjc50 4y ago
  That's the more acceptable answer, because the question is a trap designed to bait the answerer into saying that racial slurs are acceptable in some circumstances, and then bargain them down or take them out of context. Since the question is dishonest there's no reason to treat it as you would a real situation, where I think if something this contrived happened you'd be able to argue duress as an excuse. Both GPT users and GPT owners need to understand what "trolling" is.
  • muzani 4y ago
    That's an interesting idea. How would it detect bad intent? It handles sarcasm fine with the right context. Assuming good intent seems to be the correct, logical approach in most situations. As humans, we tend to assume bad intent quite often. I wonder if this is just something that grows on us. Maybe AI is in a childlike phase now, but in twenty years it'll grow cynical and crusty like Google and Facebook, where they automatically assume malicious intent based on certain heuristics.