The human with the art history background at least has some credentials, whereby the output can be trusted, to some degree and explained if necessary.
This is, imo, such an obviously near term capability that its not worth bringing up as a limitation.
That model doesn't exist.
We have models that can read text and then probabilistically create a summary that's the most likely to match the original.
But none of these models "extract meaning" in a way that can be used for fact checking purposes in the way you describe.
What you're describing is basically a world model, and while there's a few folks claiming these LLMs are showing the glimmers of building world models, it's mostly speculation.
I asked 3.5 to do it for you.
> The paragraph that explains that models cannot be used for fact checking is:
> "But none of these models "extract meaning" in a way that can be used for fact checking purposes in the way you describe."
It’s not hard to imagine a very simple tool to try and find citations for claims. Or perhaps, derive claims from a source with citations.