Post #2673291

2026-05-12 09:56 UTC

Software has been "built by machines, directed by people" for decades. That's what compilers and linkers do, that's what uncountable lines of Bash and endless CI/CD pipelines are – machines building software, directed by people. And for decades, the bottleneck has not been churning out code. It was code review, it was quality control, it was bug fixing. AI slop makes that *worse*, not better: https://freakonometrics.hypotheses.org/89367 GitLab, and the rest of the industry, is solving for the wrong problem.

Replies (8)

@rysiek@mstdn.social 2026-05-12 10:01

The outcome of this will be *worse* software. It will feel even more plastic, it will be even more brittle, and it's not because we don't know how to write better, more reliable software, but because the industry decided that writing better software is not how money is made. And GitLab just went all-in, announced to the world: we're here for plastic software, we're here for shit quality code, we're here for forcing people to review unreviewable slop and then blaming them for the bugs.

Open ##2673292
@janneke@todon.nl 2026-05-12 09:59

@rysiek@mstdn.social Cc: @regtur@mastodon.social @civodul@toot.aquilenet.fr

Open ##2673299
@lazyb0y@mastodon.social 2026-05-12 10:43

@rysiek@mstdn.social they try to handle this by using multiple Agents now, and one gets the „personality“ of the reviewer, one for writing the concepts and requirements etc. but if no human cares to thoroughly sit down and describe precisely what the real needs are that the software and its features must fulfill and how the fulfillment of these requirements can be verified (acceptance criteria) , a project will never fulfill the needs of humans or the company they work for

Open ##2673320
@tobiaspatton@cosocial.ca 2026-05-12 11:16

@rysiek@mstdn.social loved this article. Very well thought out.

Open ##2673324
@maaikees@hachyderm.io 2026-05-12 13:49

@rysiek@mstdn.social I am so glad I came to this article thanks to you! It manages to clearly state the worries I have as a test consultant in a world where no one seems to care about quality anymore, in a way I could never find the words for. I also learned about the term credence goods for the first time.

Open ##2673325
@mathaetaes@infosec.exchange 2026-05-12 18:42

@rysiek@mstdn.social Gitlab isn't solving a problem. They're optimizing based on near-term incentives. Right now, the investment money is all chummed with AI hype. If you want to see near-term valuation increase, you pivot to AI. It's just like in the '90s when anyone could say 'web' in a VC pitch and get funded. ... And just like in the early 00s, when that bubble bursts, some companies will have figured it out, and a lot of companies will fail catastrophically... and there will be people who redirected their entire career to align with what industry was asking for, who will left with difficult-to-market skills. And nothing is going to change, because every VC sees their investment as the next Amazon rather than the next Pets.com.

Open ##2673327
@RedMacryon@pounced-on.me 2026-05-13 05:22

@rysiek@mstdn.social They are solving for profit which as usual leads to an objectively worse outcome. I am as usual not surprised that they are doing this unfortunately

Open ##2673329
@octade@soc.octade.net 2026-05-13 06:07

AI is the new tulip mania gone to seed.

Open ##2673330