I’ve been wrestling with this in my head for a few days and can’t put a solid number to it, so I’m going to ask Bluesky instead. What number would you assign to the probability that in five years AI will be good enough that people will be paying to hire drop-in AI knowledge workers?