I'm one of the early predictors of AGI, and I still don't expect a rapid takeoff - even if we do something in a lab, it doesn't mean broad adoption has been achieved and further the benefits from creating such things have to cycle through the economy. I will note though that modern predictions have been consistently more pessimistic than results (see ML surveys by Bostrom et al, or various predicted benchmarks such as the big MATH dataset miss).

​

This all said, earlier responses to you are right - the modern take on education is unhealthy. It used to be acknowledge that an educated populace was a public good in and of itself. Continued learning should be undertaken simply to improve yourself (and the world around you).