5 ms·

I think everyone should now make mistakes so we ca distinguish human vs ai.

• zeristor 6mo ago
  This can be optimised for no doubt, adversarial training is like that