It’s really easy to learn how to use software assistants. Those engineers can hop on or off any time. It’s really hard to build up fundamental engineering skills.
It’s really easy to learn how to use software assistants. Those engineers can hop on or off any time. It’s really hard to build up fundamental engineering skills.
Anyway, like any analogy, it has its limits, and trying to base the whole discussion in that analogy is not productive
If there is little to no professional work in a scope, people don't get smarter, they get dumber, and it happens through simple atrophy.
Good engineering is the constant mental build up, natural tear down (atrophy) involved, where reality favors us with resiliency.
AI replacing large swathes of work will cause those neurons to not fire as often, eventually you'll lose the connections.
I don't think there is any safe amount of use that doesn't already create a persistent, self-fulfilling cycles of demand, where overuse would naturally occur just to compete in the labor market.
Its effectively a n-body problem with no visibility since the measurements would ultimately be based on value which is subjective. It collapses the value of inherent skilled labor down to 0 paving the way for slavery or death when natural dynamics collapse the economic production system.
I somewhat disagree with this statement. Just look in this thread and you will see people who have tried using AI assistants and got no utility out of them. Prompting is a skill. Here are some areas that a skilled user of AI needs to develop:
* What sort of questions are good for AI to answer
* What context/information is needed for an AI to give a good solution.
* Identify when the AI is hallucinating and going in loops.
* Which of the many AI tools to use in a given situation.
However I agree that engineers still need to build fundamental engineering skills. I am not sure what the next generation will do, but in general I think the kids will be alright.I jest, of course, but I have very specialized skills in an industry that requires esoteric bespoke things, so I feel less at immediate risk. That said, I should start building those skills, and I do see a lot of older engineers being left in the dust if they don’t adapt. Especially web / front end dev and dev ops, any field that has a bunch of boilerplate and repeated conventions will suffer first.
Imagine if it was driving a truck instead. Anyone, realistically, could've learnt to do it. But if you don't already know how to drive one, you're never going to get hired as one. And by the time the truck driving jobs are the only jobs there is left, it's a tad late - you're already behind everyone else who's got some years of experience.