I asked Claude-2 to summarize Jeremy Howard's recent essay on the danger of consolidation in AI (see https://www.fast.ai/posts/2023-11-07-dislightenment.html). • Claude's summary includes, "More oversight and regulation of AI systems is needed to ensure safety and prevent abuse." • That's precisely the point of the essay (large orgs will use regulation to keep AI for themselves) but Claude wrote the opposite. • Claude-2 is interesting, but be careful.