What actually mattered was not the profession of the user, but their expertise. The more domain experience someone had, the more successful they were in using Claude Code in that domain. And, even more interestingly, the more useful output they got from Claude from each prompt.

This is because each session begins at the center of the manifold, and human expertise is required to push it further and further into distant regions. The more expertise you have, the more you can push the session in a specific direction, the closer you get to uncovering something genuinely useful in the LLM’s latent space. Interacting with a chatbot is sort of like an exercise in highly selective decompression.