AI is the most accessible way to do a deep dive to investigate a story. AI systems provide a range of references to related work, both alternative current sources and historical context sources.

But the vast majority of news readers only want headlines and bias confirmation.

AI chat provides a natural way to follow a thread of interest via follow up questions.

I simply don't trust such systems to be accurate, reliable, or complete enough to use for that sort of task.

How is it the most accessible way when stories are paywalled?

What good is a "range of references" if the higher quality references are not accessible to the AI?

You also claim readers want "want headlines and bias confirmation" but AI is very well known to engage in sycophancy.

AI will spew complete bullshit the moment it does not have data. It is about the worst place to go to investigate a new story I don't know much about.