Are #AI-generated summaries suitable for studying and research?
tl;dr: no, they are not.
First, because they are simply not good at it, with testing indicating the latest models achieve at most 66% accuracy, with a tendency to over-generalize findings.
Second, and more importantly, offloading summarization to an AI robs researchers of an important step in the learning process. Struggling with the literature makes us better at what we do - that's the point, the summaries we create are just the residue of that process.