[ACCount37]

> Are these companies slowly gaining rot in those codebases?

Yes!

And that "yes" holds regardless of whether they use AI or not.

Let's not pretend tech debt accumulation is somehow an AI problem. Some of the world's biggest companies routinely ship code that reeks of years of rot and decay.

[jason_oster]

It's safe to say that most people romanticize their past experience. Anyone's "golden age" is someone else's "end times".

Software has always sucked [1]. "Software crisis" was coined in 1968. There has been no period of time since that it got any better. Trauma just doesn't carry the same way that triumphs do.

Perhaps people do not actually learn from their mistakes after all.

[1]: https://www.stilldrinking.org/programming-sucks

[joaquieneCnix]

> Perhaps people do not actually learn from their mistakes after all.

which is the way the economy runs. cascades of PR narratives and economic measures carrying over upscaled patterns of the very same causal chain that makes people forget to self-debrief.

but most people enjoy being leisurely ignorant whenever they find themselves immersed in such qualia so even the brightest just smile while standing by ...

as long as we read, we will get hyped one way or another.

the issue seems that some narratives get disproportionate "air time" ... which, IMO, is exactly because there are so many bystanders among the capable & resourceful ...

[lelanthran]

> Let's not pretend tech debt accumulation is somehow an AI problem. Some of the world's biggest companies routinely ship code that reeks of years of rot and decay.

While that is true, there's a large difference you are ignoring: a company can survive this type of rot indefinitely if it takes place at a certain rate. Their customers get time to acclimatise , the processes take into account the breakage, and adapt to compensate, etc.

If the AI boosters are to be believed (2x, or 10x or even, in some cases, devs claiming 100x productivity), stuffing 2 decades of rot into the next two years could prove fatal.

[gnatolf]

I don't disagree, but AIs 'willingness' to do the most painful chores mitigates a lot, I think

[layer8]

This is usually due to time pressure and cost. AI arguably substantially reduces that kind of pressure.

I’m overall skeptical as well, but the argument about tech debt is more nuanced than your comment implies.

[lenkite]

Our production incidents have dramatically increased over last 2 years after management pushed AI. In 2023, we never failed our SLA. Now, we continuously keep failing it due to some stupid outage. There was no real change in architecture or strategy - only the adoption of AI.

There have already been several high-profile AI-root-cause incidents at many companies this year. I guess we can look forward to even more deterioration in the next 2 years.

[asdf88990]

> This is usually due to time pressure and cost. AI arguably substantially reduces that kind of pressure.

This is categorically false because time and cost pressure is relative and if anything AI has exacerbated the exceptions for ROI and Turn Around time.

This along with the fact that even state of the art AI tends to generate subpar (albeit working) code extremely fast and in large volume.

You can see the effect of this already in the increased rate of production incidents recently.

This isn’t to say that AI is useless, but the tech debt it generates is unprecedented.

[rudiksz]

AI easily generates PR's with thousands of lines. I my company there's even a directive for agents to split up their work into multiple PR's based on some arbitrary metric of files/lines changed. When one of my colleagues uses AI to "implement" something, the amount of code I need to review is much much more. If I were to do review as I am supposed to, then AI is absolutely increasing the time and effort I have to put into a change (as opposed to the person who just writes a prompts and calls it a huge productivity boost).

There's the meme in programming that goes something like this "Give a programmer two lines of code to review and he'll find 20 issues. Give a programmer 500 lines to review and he's say that it looks good". At some point you simply cannot keep pace with the amount of generated code as a human.