Discussion about this post

User's avatar
Science: Outside The System's avatar

I agree that industry needs to better define what careers and jobs are needed, and that the need for new types of unique jobs evolves faster than the government and education system's ability to recognize them. Putting employers in charge would help speed up the process of educational training for new jobs. Another thing I notice is that this is a counterargument to the idea that AI will destroy jobs, since, as you imply above, it likely will create new types of jobs at least as quickly as it destroys the need for others. Indeed, we are seeing the new types of jobs becoming prominent during the AI era that are neither blue collar nor white collar, but somewhere in between. Skilled trades like being an electrician were always sort of a hybrid job, since it requires math in addition to physical skill, but AI is taking that even further. There are even jobs that require the wiring ability of an electrician and the coding ability of a software developer.

Blockchain technology has the potential to do the same thing, as there are people who were originally software developers that have gone on to start their own businesses using bitcoin mining to heat water, hot tubs, and greenhouses. This requires the skills of both a software developer, since you need to know code, the skills of a hardware engineer, since they need to know what kind of computer hardware is needed, and even the skills of an HV/AC technician for safety reasons like preventing legionnaire's disease bacteria from growing in the water. It actually does not make sense to think of jobs as white-collar and blue-collar anymore.

Skivverus's avatar

"Learn useful underlying skills" is I suppose less of a slogan than "prep for college" or "prep for a career", but I think that's what both slogans are gesturing at, and which the first used to be a good proxy for. Goodhart's Law kicked in, because it turns out lots of schools and colleges can coast for quite a while on reputations and selection effects without actually teaching all that much.

Goodhart's Law will probably kick in eventually for this too, but that doesn't mean we can't get a good few decades out of it before moving to the next measure if we do it right.

6 more comments...

No posts

Ready for more?