Comment by XorNot
9 days ago
All of those professions you've listed require about half a decade of dedicated training to be legally allowed to practice. For example an electrician takes like 7 years to become qualified, that's a full time apprenticeship, and it pays badly in the meantime.
A fact endlessly annoying to electrical engineers who legally can design their houses power system but not work on it.
(I mean I think a barber is quicker, but one of that list is also not like the others)
There are a lot of code issues you can run into when actually installing and repairing wiring that your average EE wouldn’t know much about at all. And just because you can design the power system for a building doesn’t mean you have any how to fish romex through a wall.
That’s like saying a mechanical engineer should necessarily be able to work as a machinist. Some of them can for sure, but it’s not something they are required to learn.
Also in many places anyone can do electrical work on their own house.
I'm not trying to say 'everything is fine, nothing bad is happening a world of recurring technology and industrial revolutions.' It's not. The way things are set up is bad.
My point is the author writes a column about how GPTs are ruining the ability for people to make scalable products, because when everyone can make one, nobody cares... my point is that that's not the result of GPTs. It's a result of survivorship bias skewing how we look at things.
When your business is a flywheel than needs to be running to provide a benefit to each user, then getting that flywheel running is a huge problem. The vast majority of non-scalable businesses, almost by definition, provide each customer with a benefit regardless of whether anyone else uses it. That is how you create basic, word-of-mouth, free "earned" marketing.
Half a decade is over in a flash. The current software engineering downturn is almost 3 years old now.
> For example an electrician takes like 7 years to become qualified
Using a similar playbook for what happened to SE (loads of good ~free info, bootcampification) you can drastically reduce the time to be economically useful, but that can only go for so long due to saturation. Especially residential electric work (or other trades) is no witchcraft that would require anywhere near several years of education/practice. And yeah, the legal system would have to cooperate, which will only happen when the pressure becomes high enough (see e.g. debates around "Meisterpflicht" in Germany).
That said if AI continues to progress like it currently does we may not get to that part.