After AI, what work is left that’s meaningful and pays?

I read this NYT piece and it pushed on the part that makes me uneasy: we always talk about “jobs lost,” but not what the next stable thing is after the churn.

https://www.nytimes.com/2026/07/13/business/economists-ai-threat-jobs.html?unlocked_article_code=1.xVA.8iwD.8XPZrSpF01m9&smid=bs-share

Creative destruction is normal, sure, but AI feels like it eats both the repetitive work and the “nice office job” middle layer at the same time. In architecture school I keep thinking the only stuff that stays is messy physical-world work: sites, permits, clients, repairs, care work, anything where the constraints are real and constantly changing. What do you think is the next economically viable lane that’s still human-sized, not just “own the robots”?