> Based on my experience, that will eat all software and software-related meetings. Once lab researchers focus on fields other than software, they will eat those nearly as quickly.

I don't understand your framing, because this is far more apocalyptic than any old bubble burst, even if the bubble is a massive one. A crash due to a bubble causes a temporary setback and readjustment that slowly rebounds over time. Fully displacing even 20% of the workforce would likely cause a societal implosion on a never before seen scale. Now imagine how bad it would be if it's 30, 40 or 50% of the workforce. Full automation is not temporary, it would remove these jobs for good. There is no rebound to be had, and only a mass restructuring of society that we all know is not likely to happen would save us from near-total collapse. I don't think AI labs have much of a reason to pace themselves to prevent this, because they're the ones who will come out on top in that poorer, more desperate, more divided hypothetical future.