I’m convinced that some people are overly susceptible to the “preference optimized” nature of AI output and end up completely blind to its quality and usefulness.

Not to say it’s useless garbage, there is some value for sure, but it’s nowhere near as good as some people represent it to be. It’s not an original observation, but people end up in a “folie a deux” with a chatbot and churn out a bunch of mediocre stuff while imagining they’re breaking new ground and doing some amazing thing.