Interesting, thanks!. I see this as falling into the bucket of taking unstructured data/thoughts and mapping it well to structured/articulated data that is alre

Interesting, thanks!. I see this as falling into the bucket of taking unstructured data/thoughts and mapping it well to structured/articulated data that is already out there. LLMs do excel at this and keep getting better fast (minus the hallucination, but that is easy to cross check). Would love to see examples of truly novel thesis generations that are verifiably plausible, which aren't a remix of what came earlier...