3 ms·
2. - My impression is that a lot of "AI"'s and chatbots have coped quite poorly when interacting with malicious users.
by bell-cot 4y ago
2. - My impression is that a lot of "AI"'s and chatbots have coped quite poorly when interacting with malicious users.
- saliagato 4y agoI don't think I get what you are referring to with malicious intent, but OpenAI has some "Safety best practices" [1] that don't seem bad. I'm also sure that they will improve over time. [1] https://beta.openai.com/docs/usage-policies/safety-best-practices https://beta.openai.com/docs/usage-policies/safety-best-prac...