Paolo - I read your article, but I disagree with a few fundamental assumptions.
1. There is an implicit assumption that small changes in product behavior is a small change in the underlying code.
2. AI will increase productivity by a huge amount for production ready code.
3. Developers spend all the time coding.
I will take the last one first. When I was managing large teams, my assumption is that a developer spends about 25% of the time actually coding. Let us say, AI makes them 50% more productive. So, you got a real gain of 12.5%. It is nowhere as huge as you put on your chart.
AI coding seems really great for one-off prototypes or some small, well-defined pieces, but they are not ready for production code. Multiple research papers have found AI code to be less secure, more buggy etc and they hinder rather than help experienced developers. So, you are going to lose even more of the hypothetical 12.5% gain here.
Finally, your post seems to say that the developers are not prioritizing your work, not that they are not working. Even if they are more productive, there is still no guarantee your feature will get prioritized.
The solution to your stated problem seem to be inability to get your work prioritized and nothing to do with AI or developer productivity.
paolop · 4m ago
Thanks for reading, vaidhy!
Not sure those assumptions are in there to be honest. The chart is simply to
compare replacing "devs" with "devs + AI" to obtain the same output vs. keeping the same number of devs and increase the overall output. We can debate the multiplier, but it's math that with the same multiplier, more devs will produce more output, all things being equal.
If what you are saying it's true, i.e. AI is borderline useless, it should not be used as a reason to justify layoffs, right?
Finally, I was lucky enough to get a lot of my work prioritized - that paragraph was just trying to be the usual funny interaction between PMs and devs. I guess it didn't work. The main point remains - products today are not that polished or complete, so clearly some code is not being written or fixed.
duxup · 36m ago
Every story I read about this happening is like a LinkedIn fairy tale without any actual details.
I feel like executives find it useful to pad their resume / PR and signal how “with it” they are even if the layoffs are really about something else.
paolop · 38m ago
Are companies using AI just to justify trimming the fat after years of overhiring and allowing Hooli-style jobs for people like Big Head? Otherwise, I feel like I’m missing something—why lay off developers now, just as AI is finally making them more productive, with so much software still needing to be maintained, improved, and rebuilt?
techpineapple · 31m ago
Theory --
A VP's job isn't hard because they have to manage 250 developers, A VP's job is hard because they have to manage the cognitive load of a vision for the equivalent of 250 developers worth of work. Same sort of idea with Product Manager's, Manager's, etc. If those developers are more productive, than great, you can do the same thing with more developers, but it's really hard to actually increase the number of products/business units etc. that you then have to market, PM/etc.
As for why not just increase quality? Probably because of diminishing returns.
almosthere · 41m ago
The vast majority of tech layoffs are driven by outsourcing to other countries, not to AI.
I will take the last one first. When I was managing large teams, my assumption is that a developer spends about 25% of the time actually coding. Let us say, AI makes them 50% more productive. So, you got a real gain of 12.5%. It is nowhere as huge as you put on your chart.
AI coding seems really great for one-off prototypes or some small, well-defined pieces, but they are not ready for production code. Multiple research papers have found AI code to be less secure, more buggy etc and they hinder rather than help experienced developers. So, you are going to lose even more of the hypothetical 12.5% gain here.
Finally, your post seems to say that the developers are not prioritizing your work, not that they are not working. Even if they are more productive, there is still no guarantee your feature will get prioritized.
The solution to your stated problem seem to be inability to get your work prioritized and nothing to do with AI or developer productivity.
If what you are saying it's true, i.e. AI is borderline useless, it should not be used as a reason to justify layoffs, right?
Finally, I was lucky enough to get a lot of my work prioritized - that paragraph was just trying to be the usual funny interaction between PMs and devs. I guess it didn't work. The main point remains - products today are not that polished or complete, so clearly some code is not being written or fixed.
I feel like executives find it useful to pad their resume / PR and signal how “with it” they are even if the layoffs are really about something else.
A VP's job isn't hard because they have to manage 250 developers, A VP's job is hard because they have to manage the cognitive load of a vision for the equivalent of 250 developers worth of work. Same sort of idea with Product Manager's, Manager's, etc. If those developers are more productive, than great, you can do the same thing with more developers, but it's really hard to actually increase the number of products/business units etc. that you then have to market, PM/etc.
As for why not just increase quality? Probably because of diminishing returns.
No comments yet