3 ms·
I wonder how OpenAI etc models would perform if the user says they are working for the Iranian government or something like that. Or espousing illiberal / anti-
by HPsquared 1y ago
I wonder how OpenAI etc models would perform if the user says they are working for the Iranian government or something like that. Or espousing illiberal / anti-democratic views.
- charlieyu1 1y agoThe proper thing to do is to either reject due to safety requirements or do it with no difference.
- HPsquared 1y agoIn theory, yes