I'm a software engineer, completing 10 years of professional experience this year. I started my career as a web frontend engineer (it was easier for me to de...
If the only problem is that your code is slop and nobody can work with it without AI, then it’s probably not that bad. Text models I can run locally on my five year old Macbook are maybe a year behind in terms of coding assistance. So AI for coding is probably never going away. The worst case for someone in this scenario is just that it gets a bit slower and dumber and that they have to hire more engineers again. It’ll suck but I think it’s survivable. Someone would have to make a new Stackoverflow though if we’re going to google stuff again.
Now if you integrated multiple AI services into all your business workflows and into the products you sell, on the other hand, that might be a different story. In a way the risk is the same as with cloud providers. You get locked into a stack and then your product literally dies if the provider decides you’re not paying enough, because you have no feasible way out. Tbh I would much prefer working at a post-bubble era software company fixing the codebase to working at a random company now extracting their IT from a hyperscale cloud. But in reality, most companies that bet on AI are in this scenario. Nobody only installed Claude and called it a day.
I’m sure it will get there, but the biggest issue i see with the current models and implementations is that without good, or in some cases excellent, guidance, most can’t handle anywhere near a full codebase and even the best models are happy to take your question about fixing any particular task you throw at them back to square one with no regard towards how the existing platform is set up. Even for a relatively light codebase that it is currently building, it’s happy to scrap it start over in order to solve whatever potentially poorly worded issue you’re trying to track down. And that’s not even taking into account how quickly it can go from trying to solve an issue to nerfing tests in order to pass. Frustrated that things are no longer passing an expected integer test? Easily solved by allowing floats or nulls to count as passes, and damn if that isn’t a process to correct three issues deep after you’ve ‘validated’ that issue and it’s now being discovered down the line.
If the only problem is that your code is slop and nobody can work with it without AI, then it’s probably not that bad. Text models I can run locally on my five year old Macbook are maybe a year behind in terms of coding assistance. So AI for coding is probably never going away. The worst case for someone in this scenario is just that it gets a bit slower and dumber and that they have to hire more engineers again. It’ll suck but I think it’s survivable. Someone would have to make a new Stackoverflow though if we’re going to google stuff again.
Now if you integrated multiple AI services into all your business workflows and into the products you sell, on the other hand, that might be a different story. In a way the risk is the same as with cloud providers. You get locked into a stack and then your product literally dies if the provider decides you’re not paying enough, because you have no feasible way out. Tbh I would much prefer working at a post-bubble era software company fixing the codebase to working at a random company now extracting their IT from a hyperscale cloud. But in reality, most companies that bet on AI are in this scenario. Nobody only installed Claude and called it a day.
I’m sure it will get there, but the biggest issue i see with the current models and implementations is that without good, or in some cases excellent, guidance, most can’t handle anywhere near a full codebase and even the best models are happy to take your question about fixing any particular task you throw at them back to square one with no regard towards how the existing platform is set up. Even for a relatively light codebase that it is currently building, it’s happy to scrap it start over in order to solve whatever potentially poorly worded issue you’re trying to track down. And that’s not even taking into account how quickly it can go from trying to solve an issue to nerfing tests in order to pass. Frustrated that things are no longer passing an expected integer test? Easily solved by allowing floats or nulls to count as passes, and damn if that isn’t a process to correct three issues deep after you’ve ‘validated’ that issue and it’s now being discovered down the line.