Let’s assume that in 10 years, AI has advanced absurdly, insanely fast, and is now capable of doing everything a Senior SWE can do. It can program in 15 different languages, 95% accuracy with almost no mistakes, can create entire applications in minutes, and no more engineers or SWEs are needed… What will all the devs do? Do they just become homeless? Transition to medical field, nursing? Become tradespeople like plumbers, HVAC?
But I mean many people have already lost their job because AI automated it away.
True, and many people have lost jobs because something else automated it away, like toll booth workers, grocery clerks, and telephone switchers, and computers (i.e. people who would compute things by hand).
Jobs disappearing because technology advances is natural. It sucks for those impacted, but it’s natural, and IMO it’s only a problem of new jobs aren’t created fast enough, or whole industries disappear. Fighting to keep jobs in spite of automation runs the risk of having an entire industry disappear, such as if dock workers win the fight to prevent automation on the docks, they’ll just all lose their jobs at the same time once automation can replace them all at once.
The better plan is to adjust and adapt as technology changes. If you’re entering CS or a recent grad, make sure you understand concepts and focus less on syntax. If you’re a mid level, learn to incorporate AI into your workflow to improve productivity. If you’re a senior, work toward becoming an architect and understand how to mitigate risks with poor quality code.
Fighting AI will at best delay things.
That is certainly true. It just sucks, that so many people are scrambling for jobs and rich people get richer. There has to be a better way.
I think the root is that change is scary and hard. As a dev, I feel that too, but I also think people are overreacting a bit.
Just like how COBOL still exists, traditional software jobs will continue to exist, it’ll just be harder and harder to find those roles as companies find more and more use cases for AI. But it’ll also take several years for companies to get on board. So I’m not too worried, though I can’t recommend my field to college students unless they’re really interested, because it could be a bumpy ride.
For me the other scary thing is the loss of control. We are already drowning in code everyone needs and noone understands and now we build systems, who can produce mountains of that.
But then again you could use it to explain code. We will so quickly become dependent on it.
This seems to assume that human written code is better than AI written code. That’s currently case, generally speaking, so we need plenty of checks in place. But once it catches up, is there really a difference if an AI or human writes the code?
I don’t look forward to debugging generated code, but that’s assuming the current state of code gen. In hypothetical future where most code is AI generated, surely we’ll expect readable or provably correct code.
I’m not so much worried about the functionality of the code, but rather the maintainability. A lot of code was written, by a person long ago, copied together from forums online refactored by people, who didn’t understand it, but at least at some point someone understood it good enough to create it.
I find the idea worrisome, that we will deal with code no one ever understood, it’s just kind of there and seems to do what we want, and now you have to change it.