First, I wouldn't put it past OpenAI to charge even if your prompt triggered the guardrails, transferring responsibility is their style.
Second, how does vague billing prevent someone from mapping the "I'm sorry, I can't help you with that" style of responses to map the guardrails?
Lalabadie · · focus · HN ↗
boredatoms · · focus · HN ↗
schaefer · · focus · HN ↗
And if they do that accurately, there's no stopping someone from mapping the guard rails.
Clearly that outcome is more uncomfortable than what we have now, which is enough people seem to be okay with vague billing.
ASalazarMX · · focus · HN ↗
Second, how does vague billing prevent someone from mapping the "I'm sorry, I can't help you with that" style of responses to map the guardrails?