All of these criticisms fail to address a big elephantine question: Are these hurdles insurmountable? If ChatGPT's model fails to differentiate or understand t

All of these criticisms fail to address a big elephantine question: Are these hurdles insurmountable? If ChatGPT's model fails to differentiate or understand the question of credibility, how tough would it be to build classifiers that could do a 75% decent job? If AI presentation of knowledge is imperfect - well, fine. But is human presentation or selection perfect for that matter? Criticising a thing for imperfection is not a viable take. If ChatGPT's knowledge is limited, how tough would it be to plug it into larger and larger datasets, including the live internet itself? It's like there's a wall between the people who see the potential of this and the people who feel so threatened by it that they refuse to look over the wall.