I think the company I work for has a pretty pragmatic approach for AI utilization. Don't use it to replace jobs, use it to make jobs more efficient (we call it an "enhancer"). IE we won't lay people off, at least not because of AI, we just may not hire as quickly as we did before LLMs, but still attaining similar scale. This goes for all departments, not just tech. We have very strict standards on what code gets accepted (no slop). Whether it's AI generated or not is irrelevant; even senior engineers can be guilty of slop after a long day. But we do have regular meetings to discuss our AI usage and to converge upon common patterns and rules to level everyone up.

That being said, I think we as a society are quickly reaching the point where there just doesn't exist enough jobs to keep everyone gainfully employed. There may be new jobs opening up but not ones that people in the middle or ends of their careers can quickly pick up on. I don't think this is a bad thing necessarily, just that when combined with the shitty safety net that the US has, it's a recipe for disaster. If even folks who are working full-time still need assistance for groceries, then we've already failed, and that's not even AIs fault.