I think everyone should now make mistakes so we ca distinguish human vs ai.
This can be optimised for no doubt, adversarial training is like that
This can be optimised for no doubt, adversarial training is like that