This is fine as far as it goes, but it shows that if you know the answer (and, in this case, have a hypothesis about why ChatGPT went astray with a more open pr

This is fine as far as it goes, but it shows that if you know the answer (and, in this case, have a hypothesis about why ChatGPT went astray with a more open prompt), you can prompt it to do better.