> some people don't like LLMs

It's not even that they "don't like LLMs". They just don't like academic fraud! If references were fabricated with a Markov chain it would be just as bad!