This is an interesting perspective. I've sort of intentionally pushed towards working where AI isn't so useful (physical workflows that are heavily dependent on

This is an interesting perspective. I've sort of intentionally pushed towards working where AI isn't so useful (physical workflows that are heavily dependent on inputs from humans—scientists in particular) but my experience nonetheless has been discovering how useless AI is in so many ways I envisioned it replacing people by now. It's extremely useful in very narrow bands of application, and outside of that, it's often more of a distraction than an advantage. I have a hard time believing it's making people that much more productive. It certainly helps me here and there with very specific low-level implementations, but the really important, higher-level work I do? The way I decide which low-level work to do in the first place? Not really, no. I have to interface with very non-technical people who need bespoke solutions to their problems. I need to tie implementations for them together with existing systems that are not standardized, not well-known, and often poorly documented. I need to consider how the life cycle of these solutions can integrate with that of others, how it fits into the workflow and capacity of myself and people I work with, etc. AI can't do any of that properly right now, and I don't expect that it will any time soon. If I tried to get it to work, I'd likely spend as much time fighting Claude as I'd save. I don't know... What are people doing that they can actually be replaced? Or that companies could decide they actually need fewer people? My suspicion is that with money being more expensive to borrow, teams are staying lean because we were absurdly inefficient as an industry for the better part of a decade. That's not an AI thing, but a staying closer to actual means thing.