8 ms·
If someone is taking the time to refute ChatGPT’s output and telling you why the answer isn’t applicable in a given situation, it certainly implies that ChatGPT
by drewbug01 11mo ago
If someone is taking the time to refute ChatGPT’s output and telling you why the answer isn’t applicable in a given situation, it certainly implies that ChatGPT wasn’t “correct enough” at all.
What situations do you think it’s fine to be “correct enough?”
- paulcole 11mo agoBut this person says they want to refute it in every situation. Some people seem to make a hobby of refuting the output of others. So no, I don’t trust the implication that if somebody spends time refuting it that it must be worth refuting. In my experience (with both people-output and ChatGPT-output) my goal is to not refute anything unless it absolutely positively must be refuted. If it’s a low-stakes situation where another person has an idea that seems like it might/will probably work, let them go nuts and give it a shot. I’ll offer quick feedback or guiding questions but I have 0 interest in refuting it even if I think there’s a chance it’ll go wrong. They can learn by doing.
- Aeolun 11mo ago> But this person says they want to refute it in every situation. No. You read that I want to refute its output in every situation. Obviously I don’t have to do that if ChatGPT is correct, but people don’t generally quote ChatGPT at you if it’s just confirmed what you already said.
- paulcole 11mo agoYes. I read that because that’s what you said: > I don’t think I’m haughty when I don’t want to try and convey 20 years of experience in a few sentences every time some quotes chatgpt? Could you explain how “every time some[one] quotes chatgpt” should be read as “not every time some[one] quotes chatgpt”?