Hacker News Reader

Comment by junon | Hacker News Reader

That's definitely happening lol

Maybe I should rephrase, it's not happening at any org that will be in business beyond the next couple years. Customers don't care about your software factory, they care about a product that works reliably.

> they care about a product that works reliably.

I actually disagree here, what matters most is procurement & distribution. You can sell absolute dog water software (see: Teams[1]) and capture a market with the right procurement and distributions pipelines.

The sad reality IMO is that the quality of software matters much less than we (the engineers) would hope it does.

[1] https://web.archive.org/web/20260322032609/https://www.theve...

I still believe that quality of software matters in the long term.

The issue with that belief is that it's basically impossible to falsify.

There's obviously some kind of quality floor, it's odd that folks are pretending otherwise.

Teams is dog water, but it's dog water that works.

It's more that Teams is already included in a bundle of actually useful and working software in the form of Microsoft Office.

Yeah I can use it to see and hear my coworkers and their screens and exchange text messages through a crappy editor and schedule meetings. That's really all a business needs to keep using it.

This is the entirely wrong framing.

The right framing is ecosystems.

> Customers don't care about your software factory, they care about a product that works reliably.

You don't think AI is capable of putting out CRUD apps into the ether that work reliably? Even my local GLM5.3-Flash is capable of doing that. These products make up a non negligible amount of the industry. Additionally reducing headcount is about to hit in the next year or so. Where are juniors gonna find jobs? Etc etc...

Rails cli could do this a decade ago, there’s no major difference here other than the number if lines dumped.

> there’s no major difference here

Sure thing, I can see any sort of serious discussion is off the table.

> You don't think AI is capable of putting out CRUD apps into the ether that work reliably?

I know it isn't. No model is capable of programming well, let alone writing good software.

Instead of just making claims back-and-forth, can this thread provide sources? I don't believe either of you without proof.

No one's going to come out and admit that they're just stamping LGTM on every pull request because they can't bear to review another net 2,000 lines of AI-generated code. The problem is that right now the tools we have for code review were designed for a world where code was cheap.

Invert it. Write code yourself, have AI do the code reviewing. You'll have a much better understanding of what's actually going on.

winrar & vlc will never be forgotten

How have these two been killed by AI?

they won't be until they are?

also zamzar

I personally rarely touch an IDE anymore. I talk to lots of people with the same experience. This is anecdotal, sure, but if you ask any software developer that isn't staunchly anti-ai or just living with their head in the sand, you'll probably get told something similar.

I find this very hard to believe, even though I use plenty of frontier AI in my dev process. I have to live in vscode to get things done because I still frequently need to look at the generated code and point out issues to get the AI past some problem it's stuck at.

Improve your workflow. I have 4 instances of intellij, 2 web storm, 2 Rustrover and two phpstorm open for WEEKS. Along with some vs code instance. I think I unminimized only one just today to check something: I was being super lazy and I just wanted to check git status without typing anything. I had to rebuild multiple images with new structure to mitigate php2shell; had to patch more than 15 servers etc just in last few days. Fork old other people's images and update them, add some features while at it. Bug fix old php sites not running on new php. And so on and so on...all done with 5/6 agents running in parallel. I have the IDEs open but minimized. Don't need them really...

i like reading the generated code as a code review, and then i can leave inline comments back to the ai, rather than trying to point the ai to the right lines via chat.

im already practiced at reviewing code through code review tools, so id rather do it that way than in vscode

Same here, I abandoned reviewing code in VSCode a long time ago, I rarely use it now, I do like reviewing code in github but always hated how slow and outdated it is, but I built my own solution to replace it (pyor.review), try it out and let me know if you find it useful, I see that we kinda share the same review process, hope it helps.

I rarely open the IDE. I have about a dozen agents at all times.

Spec -> Prompt -> Validate -> Test -> Merge Review -> Production -> Agents monitor production -> Alerts -> Roll back

This is true for a many of us. All my team's code is now AI written.

> This is anecdotal, sure, but if you ask any software developer that isn't staunchly anti-ai or just living with their head in the sand, you'll probably get told something similar.

No competent developer is just letting an LLM write code for them.

Reply on news.ycombinator.com