I think both sides seem to have the same issues with the other. One side is sceptical that the other is getting good use from LLMs, and the other suggests they're just not using it correctly.

Both sides think the other is either exaggerating or just not using the tool correctly.

What both sides should do is show evidence in the form of chat extracts or videos. There are a number from the pro-LLM side, but obviously selection bias applies here. It would be interesting if the anti-LLM side started to post more negative examples (real chat extracts or videos).