Yes. Also if the training data has inaccuracies, how do those manifest? ChatGPT being wrong could be cognitive dissonance caused by the many perspectives and l

Yes. Also if the training data has inaccuracies, how do those manifest? ChatGPT being wrong could be cognitive dissonance caused by the many perspectives and levels of correctness crammed into a single NN.