I have two ideas you might like:

1) initial signs of AI adoption harming job growth got the signal backwards: It's not that we didn't hire Jrs because we had AI now, its that we didn't hire Jrs because they used AI now, and it was hard to reconcile this with our standing hireability signals: code quality, side projects, portfolio, etc.

2) The TAM argument for AI replacing all work is not even required. Anthropic and company (and several future companies) can become fantastically wealthy beyond anyone's dreams just by following claude code subscription model to make life easier for workers, who in turn gladly pay for the subcription.

both are here with more details, apologies for links to personal (ad-free) blog

https://jodavaho.io/posts/ai-signalling.html

https://jodavaho.io/posts/ai-jobpocolypse.html

> Anthropic and company (and several future companies) can become fantastically wealthy beyond anyone's dreams just by following claude code subscription model

isn't the claude code subscription model heavily subsized and not profitable on its own?