But they have engaged with it, and made an assessment about it's current utility.

We have no reason to believe that they won't keep an eye on this.

Little to nothing about AI tools so far suggests that that one can't just as easily pick the skills later. Tools that will get "exponentially better" will almost certainly be unrecognizable to someone desperately engaging with them now, for not other reason the sake of "having 1-2 years of experience"

Someone might reasonably choose to to bet on the upside. That doesn't imply that everyone else ought to fearfully hedge.

I don't think there's such a thing as a "fast take-off" where human experience with 2026-era LLM coding remains economically relevant.

Eh, I'm not super worried. After all, Every six months or so, the latest model changes everything and the former model was complete garbage. It's not just a new model—it's a new paradigm shifting the landscape of agentic development.

Won't AI getting better also mean AI will be getting easier to learn and use?

Feels to me like there are at least one or two more paradigm shifts coming in how AI gets used which will make current tools obsolete.

As one example, I think we will eventually get GUI dashboards to manage AI agents which will be easier to use than current CLI tools.