If running your business to '90s standards is acceptable, sure, you can use AI to automate your manual processes with the same error rate and keep doing the same thing indefinitely.
But if the competitors have real software engineers and have used them to actually improve reliability, you'll be left behind.
Sorry, but that logic is pure cope.
If you want to go down the value chain, then by definition the less valuable the software is and the easier to be commoditized. The automation is not going to help just the manager-turned-vibecoder, it's also going to help professionals to create FOSS alternatives that can be robust enough.
It's not going to happen overnight, but the trend is there.
> If you want to go down the value chain, then by definition the less valuable the software is and the easier to be commoditized.
I'm not sure that holds for what we're talking about - high-value software can afford to be somewhat flaky because it delivers enough value when it works to make up for it, software that's only marginally worthwhile needs to be reliable because if it isn't then it's not worth the bother. Commoditized fields are more competitive.
> The automation is not going to help just the manager-turned-vibecoder, it's also going to help professionals to create FOSS alternatives that can be robust enough.
Not convinced. In my experience these tools don't really help with creating high-quality software. Maybe they'll get there eventually (at which point we're all out of a job), but right now they can't "hit the high notes".
> Commoditized fields are more competitive.
Doesn't that also lead to the conclusion that "software engineers" are going to lose their ability to command high salaries, if the real value is in the domain expertise and not in the ability of optimizing some part of the business process?
What software engineers are being hired to work on: