[quirino]

> As a bonus, we look forward to fewer violations (exhibit A, B, C) of our strict no LLM / no AI policy,

Hilarious how the offender on "exhibit A" [1] is the same one from the other post that made the frontpage a couple of days ago [2].

[1] https://github.com/ziglang/zig/issues/25974

[2] https://news.ycombinator.com/item?id=46039274

[dreamcompiler]

My old rule about the difference between coding and software engineering:

  For coding, "it seems to work for me" is good enough. For software engineering, it's not.

My new rule:

  For coding, you can use AI to write your code. For software engineering, you can't.

[bhouston]

> For coding, you can use AI to write your code. For software engineering, you can't.

You can 100% use AI for software engineering. Just not by itself, you need to currently be quite engaged in the process to check it and redirect it.

But AI lowers the barrier to writing code and thus it brings people will less rigour to the field and they can do a lot of damage. But it isn't significantly different than programming languages made coding more accessible than assembly language - and I am sure that this also allowed more people to cause damage.

You can use any tools you want, but you have to be rigorous about it no matter the tool.

[networked]

> For coding, you can use AI to write your code. For software engineering, you can't.

This is a pretty common sentiment. I think it equates using AI with vibe-coding, having AI write code without human review. I'd suggest amending your rule to this:

> For coding, you can use AI. For software engineering, you can't.

You can use AI in a process compatible with software engineering. Prompt it carefully to generate a draft, then have a human review and rework it as needed before committing. If the AI-written code is poorly architected or redundant, the human can use the same AI to refactor and shape it.

Now, you can say this negates the productivity gains. It will necessarily negate some. My point is that the result is comparable to human-written software (such as it is).

[jorl17]

100% this.

Just don't expect to get decent code often if you mostly rely on something like cursor's default model.

You literally get what you pay for.

[epolanski]

I absolutely don't care about how people generate code, but they are responsible for every single line they push for review or merge.

That's my policy in each of my clients and it works fine, if AI makes something simpler/faster, good for the author, but there's 0, none, excuses for pushing slop or code you haven't reviewed and tested yourself thoroughly.

If somebody thinks they can offset not just authoring or editing code, but also taking the responsibility for it and the impact it has on the whole codebase and the underlying business problem they should be jobless ASAP as they are de facto delegating the entirety of their job to a machine, they are not only providing 0 value, but negative value in fact.

[strix_varius]

Totally agree. For me, the hard part has been figuring out the distinction with junior engineers... Is this poorly thought out, inefficient solution that is 3x as long as necessary due to AI, or inexperience?

[stocksinsmocks]

Not defending him, but we were already doing this with electron apps, frameworks, libraries, and scripting languages. The only meaningful cost in most software development is labor and that’s what makes sense to optimize. I’d rather have good software, but I’ll take badly made software for free over great software that costs more than the value of the problem solved.

These discussions are always about tactics and never operations.

[strix_varius]

> I’ll take badly made software for free

No, not if I have to maintain it.

Code is liability. LLM written PRs often bring net negative value: they make the whole system larger, more brittle, and less integrated. They come at the cost of end user quality and maintainer velocity.

[stocksinsmocks]

Is that not also true of human written software that costs more per hour than the monthly cost of a coding agent? Developers are expected to ship enterprise software with defects that would land you in court if you made equivalent mistakes designing a water treatment plant or bridge.

I get the “AI sucks” argument from a programmers point of view. It’s weird looking and doesn’t care about “code smells” or about rearranging the code base’s deck chairs just the way you like. From an owner’s or client’s perspective, human programmers suck. You want big standard CRUD app? Like a baby’s first Django app? That’s going to take at least 6 months for some reason. They don’t understand your problem domain and don’t care enough to learn it. They work 15 minutes on the hour, spend 45 on social media or games, and bill you $200/hr. They “pair program” for “quality” to double their billed rate for the same product. They bill you for interns learning how to do their job on your dime. On top of that there is still a very good chance the whole project will just be a failure. Or I can pay Anthropic $20/month and text an AI requirements on my phone when I’ve got 5 minutes of down time. If it doesn’t work I just make a new one and try again. Even if progress on AI stopped today, the world is now so much better for consumers of programs. Maybe not for developers unless you’re writing the AI and getting paid in the millions. Good for them. I’m glad to see the $200/hr Stack Overflow copy and pasters go do something else.

[sidjxnbx]

I get it, but I think there’s something deeply anti human about being ok with this (not just in software). It’s similar in sentiment to how you behave when nobody is looking - a culture and society is so much better off if people take pride in their work.

Obviously there’s nuance (I’ll take slop food for starving people over a healthy meal for a limited few if we’re forced to choose), but the perverse incentives in society start to take over if we allow ourselves to be ok with slop. Continuously chasing the bottom of the barrel makes it impossible for high quality to exist for anyone except the rich.

Put another way: if we as a society said “it is illegal to make slop food”, both the poor and the rich would have easy access to healthy food. The cost here would be born by the rich, as they profit off food production and thus would profit less to keep quality high.

[stocksinsmocks]

I’m pretty sure the USSR, Cuba, and the like never succeeded doing this sort of thing, but maybe if we hit ourselves in the head with the same hammer (and sickle) just one more time it will work?

[manwe150]

Does it matter? Either way seems to just reflect badly on the junior, who needs to improve their self-review skills and knowledge

[epolanski]

It's not easy to be a junior, and we might be speaking with survivor bias, but most juniors don't end up in solid engineering teams, they are merely developers that are much cheaper and from whom you expect much less, but more often than not they are borderline left learning and figuring out things on their own. They need to luck some senior member that will nurture them and not just give them low quality work (which I admit I have done too when I had myself lots of pressure to deliver my own stuff).

Even in less desperate teams, as productivity grows with AI (mine does, even if I don't author code with it it's tremendous help in just navigating repos and connecting the dots, it saves me so much time...) the reviewing pressure increases too, and with that fatigue.

[strix_varius]

It does matter, because it's a worthwhile investment of my time to deeply review, understand, and provide feedback for the work of a junior engineer on my team. That human being can learn and grow.

It is not a worthwhile use of my time to similarly "coach" LLM slop.

The classic challenge with junior engineers is that helping them ship something is often more work than just doing it yourself. I'm willing to do that extra work for a human.

[gherkinnn]

I disagree with the new rule. The old one is fine and applies to LLMs.

Vibing and good enough is a terrible combination, as unknown elements of the system grow at a faster rate than ever.

Using LLMs while understanding every change and retaining a mental model of the system is fine.

Granted, I see vibe+ignorance way too often as it is the short-term path of least resistance in the current climate of RTO and bums in seats and grind and ever more features.

[rootlocus]

I feel like the distinction is equivalent to

    LLMs can make mistakes. Humans can't.

Humans can and do make mistakes all the time. LLMs can automate most of the boring stuff, including unit tests with 100% coverage. They can cover edge cases you ask them to and they can even come up with edge cases you may not have thought about. This leaves you to do the review.

I think think the underlying problem people have is they don't trust themselves to review code written by others as much as they trust themselves to implement the code from scratch. Realistically, a very small subset of developers do actual "engineering" to the level of NASA / aerospace. Most of us just have inflated egos.

I see no problem modelling the problem, defining the components, interfaces, APIs, data structures, algorithms and letting the LLM fill in the implementation and the testing. Well designed interfaces are easy to test anyway and you can tell at a glance if it covered the important cases. It can make mistakes, but so would I. I may overlook something when reviewing, but the same thing often happens when people work together. Personally I'd rather do architecture and review at a significantly improved speed than gloat I handcrafted each loop and branch as if that somehow makes the result safer or faster (exceptions apply, ymmv).

[krupan]

"I feel like these distinctions are equivalent to

    LLMs can make mistakes. Humans can't."

No, that's not it. The difference between humans and AI is that AI suffers no embarrassment or shame when it makes mistakes, and the humans enthusiastically using AI don't seem to either. Most humans experience a quick and viseral deterrent when they publish sloppy code and mistakes are discovered. AI, not at all. It does not immediately learn from its mistakes like most humans do.

In the rare case when there is a human that is consistently persistently confidently wrong like AI, a project can identify that person and easily stop wasting their time working with that person. With masses of people being told by the vocal AI shills how amazing AI is, projects can easily be flooded with confidently wrong aaI generated PRs.

[blibble]

> LLMs can automate most of the boring stuff, including unit tests with 100% coverage. They can cover edge cases you ask them to and they can even come up with edge cases you may not have thought about. This leaves you to do the review.

in my experience these tests don't test anything useful

you may you have 100% test coverage, but it's almost entirely useless but not testing the actual desired behaviour of the system

rather just the exact implementation

[nyrikki]

The 100% test coverage metric is far more then "entirely useless" it is typically incredibly harmful.

Brittle meaningless tests tend to lock bad decisions in, and prevent meaningful refactoring.

Bad tests simply are code debt, and dramatically increase the future cost of rework and adaptation.

[gloosx]

If unit tests are boring chores for you, or 100% coverage is somehow a goal in itself, then your understanding of quality software development is quite lacking overall. Tests are specifications: they define behavior, set boundaries, and keep the inevitable growth of complexity under control. Good tests are what keep a competent developer sane. You cannot build quality software without starting from tests. So if tests are boring you, the problem is your approach to engineering. Mature developers dont get bored chasing 100% coverage – they focus on meaningful tests that actually describe how the program is supposed to work.

[rootlocus]

> Tests are specifications: they define behavior, set boundaries, and keep the inevitable growth of complexity under control.

I set boundaries during design where I choose responsibilities, interfaces and names. Red Green Refactor is very useful for beginners who would otherwise define boundaries that are difficult to test and maintain.

I design components that are small and focused so their APIs are simple and unit tests are incredibly easy to define and implement, usually parametrized. Unit tests don't keep me "sane", they keep me sleeping well at night because designing doesn't drive me mad. They don't define how the "program" is supposed to work, they define how the unit is supposed to work. The smaller the unit the simpler the test. I hope you agree: simple is better than complex. And no, I don't subscribe to "you only need integration tests".

Otherwise, nice battery of ad hominems you managed to slip in: my understanding of quality software is lacking, my problem is my approach to engineering and I'm an immature developer. All that from "LLMs can automate most of the boring stuff, including unit tests with 100% coverage." because you can't fathom how someone can design quality software without TDD, and you can't steelman my argument (even though it's recommended in the guidelines [1]). I do review and correct the LLM output. I almost always ask it for specific test cases to be implemented. I also enjoy seeing most basic test cases and most edge cases covered. And no, I don't particularly enjoy writing factories, setups, and asserts. I'm pretty happy to review them.

[1] https://news.ycombinator.com/newsguidelines.html Please respond to the strongest plausible interpretation of what someone says, not a weaker one that's easier to criticize. Assume good faith.

[brodo]

Check out this dude: https://github.com/GhostKellz?tab=repositories

He's got like 50 repos with vibe-coded, non-working Zig and Rust projects. And he clearly manages to confuse people with it:

https://github.com/GhostKellz/zquic/issues/2

[stocksinsmocks]

I don’t think this is uncommon. At one point Lemmy was a project with thousands of stars and literally no working code until finally someone other than the owner adopted it and merged in a usable product.

[PNewling]

Wow, and if you go to their website listed in they're profile, not only do almost none of the links work, the one that did just linked out to the generic template that it was straight copied from. Wow.

[forgotpwd16]

It is questionable if they've even tried any one of them.

[port11]

Hustle hustle. I'm not disgusted by this person, but by the system that promotes or requires such behaviour.

[mikelitoris]

oh god... he has a humongous AI generated PR for julia too https://github.com/tshort/StaticCompiler.jl/pull/180

[sundarurfriend]

More context/discussion on this: https://discourse.julialang.org/t/ai-generated-enhancements-...

(Honestly, that's a lot more patience than I'd be able to give what are mostly AI-generated replies, so kudos to these folk.)

[nake89]

When confronted about LLM writing completely broken tests the guy said the funniest thing: "It knows what it’s doing but tends to be… lazy."

I'm a big fan of LLMs but this guy is just a joke. He understand nothing of the code the LLM generates. He says things like "The LLM knows".

He is not going to convince anybody to merge is PRs, since he is not even checking that the tests the LLM generates are correct. It's a joke.

[Lyngbakr]

In a submission to OCaml, when asked why the files he submitted list someone else as an author he says,

    > Beats me. AI decided to do so and I didn't question it.°

I find that sort of attitude terrifying.

° https://github.com/ocaml/ocaml/pull/14369#issuecomment-35573...

[poulpy123]

I cannot believe it's not trolling

[victorbjorklund]

Yeah, either this guy's totally insane or it could even be somebody who's an AI skeptic who's just flooding projects with really dumb PRs just to show the risks and get people skeptical about the use of AI in open source (Takes on my folie hat)

[franktankbank]

Intentional or not its an interesting social experiment.

[monegator]

that's a grifter doing grifting. there was a thread on /g/ about this guy the other day, anons digged out much of its past as a failure / grifter in many areas, running away with the money at the first problem

[zipy124]

I'd be willing to put money that before LLMs they were all in ok crypto

[mzajc]

  function estimate_method_targets(func_name::Symbol, types::Tuple)
      # Conservative estimate
      # In a real implementation, we'd query the method table
      return 2  # Assume multiple possibilities
  end

Hilarious. Was this model trained on XKCD [0] by any chance?

[0]: https://xkcd.com/221/

[Archit3ch]

Among all the other problems with this... They describe [1] their contributions as "steering the AI" and "keeping it honest", which evidently they did not.

[1] https://discourse.julialang.org/t/ai-generated-enhancements-...

[sundarurfriend]

As an aside, he originally titled the thread "A complete guide to building static binaries with Julia (updated for 1.12)", with no mention of AI. That got me excited every time I opened the Discourse, until I remembered it was this slop. :/

[legobmw99]

Similar things on the OCaml forums. He has a post titled “Dune: concurrent builds are here!” but really it’s a +967 -100 slop PR that had to be closed

[conartist6]

Maybe this guy is it: the actual worst coder in the world

[CoastalCoder]

Well that's the origin story for the main character on Solo Leveling, so...

Actually, I probably shouldn't make this comment publicly. It could cause another 3-5 programmer-isekai anime series.

[smartmic]

A question I only dare to ask myself in these times of LLM: Is this even a real human being or already an instance of an ‘agentic system’?

[testdelacc1]

Lot of people are criticising this guy but we all benefit from having an example to show people - this, please don’t do what this guy is doing. Please read the generated code, understand it, edit it and then submit it.

If anyone’s answer to “why does your PR do this” is “I don’t know, the AI did it and I didn’t question it” then they need a time out.

[zipy124]

I guess we now have the equivalent of cowboy builders but for software now. Except no one asked for anything to be built in this case lol.

[ljm]

The people of Jonestown collectively drank less kool-aid than all this.

I don't know whether to be worried or impressed.

[joelreymont]

I had $1000 in Claude credits and went to town.

Yes, I made mistakes along the way.

[forgotpwd16]

The biggest mistake, AI or not, is dropping a 10K+ PR. 300~500 LOC is how far one should be going, unless they're doing some automated refactoring. E.g. formatting the entire StaticCompiler.jl source. This should've been a distinct PR, preferably by a maintainer.

[VMG]

I've seen this in other places as well.

The bottleneck is not coding or creating a PR, the bottleneck is the review.

[joelreymont]

This ought to be automated using AI.

It could first judge whether the PR is frivolous, then try to review it, then flag a human if necessary.

The problem is that Github, or whatever system hosts the process, should actively prevent projects from being DDOS-ed with PR reviews since using AI costs real money.

[messe]

> This ought to be automated using AI.

When the world is telling you to fucking stop, maybe take a moment and listen.

[VMG]

I mean this with all sincerity, try doing this yourself.

The established projects are resistant to YOLOing their projects and running them on complete LLM autopilot.

You are proposing a completely different development style.

Fork Ocaml to Oca-LLM and Julia to Jul-AI and see how it goes.

[joelreymont]

I agree with you.

[zipy124]

Please don't tell me you actually spent $1000 on generating fake tests....

[Anon1096]

It's truly astonishing to me that your account has existed since 2008 and you decided to pull this.

As a troll job for the lulz it is some amazing work. Hats off

[globnomulous]

This isn't just "making mistakes." It's so profoundly obnoxious that I can't imagine what you've actually been doing during your apparently 30 years of experience as a software developer, such that you somehow didn't understand, or don't, why submitting these PRs is completely unacceptable.

The breezy "challenge me on this" and "it's just a proof of concept" remarks are infuriating. Pull requests are not conversation starters. They aren't for promoting something you think people should think about. The self-absorption and self-indulgence beggar belief.

Your homepage repeatedly says you're open to work and want someone to hire you. I can't imagine anybody looking at those PRs or your behavior in the discussions and concluding that you'd be a good addition to a team.

The cluelessness is mind-boggling.

It's so bad that I'm inclined to wonder whether you really are human -- or whether you're someone's stealthy, dishonest LLM experiment.

[messe]

You've wasted other peoples time and mental energy with utter bullshit that you weren't even bothered to read yourself. Be more considerate in future.

[carlmr]

I'm not sure if this is advanced trolling at this point.

[amoss]

This is redefining the cutting edge of trolling.

[CoastalCoder]

I think the term is "frontier trolling".

[trueno]

trollblazing

[Sammi]

I'll one up you: at this point I'm becoming pretty sure that this is a person who actually hates LLMs, who is trying to poison the well by trying to give other people reasons to hate LLMs too.

[kibwen]

I envy your optimism. The truth is that humans are generally stupider and more craven than you have apparently even begun to conceive.

[cyanydeez]

Is the. AI bubble just biolliinaires larping about their favorite dystopuan scifi?

[rdtsc]

My favorite of his https://x.com/joelreymont/status/1990981118783352952

> Claude discovered a bug in the Zig compiler and is in the process of fixing it!

...a few minutes later...

https://github.com/ziglang/zig/pull/25974

I can see a future job interview scenario:

- "What would you say is your biggest professional accomplishment, Joel?"

- "Well, I almost single-highhandedly drove Zig away from Github"

[ivanjermakov]

> Well, I almost single-highhandedly drove Zig away from Github

If you think about it, Joel is net positive to Zig and its community!

[ljm]

Those overly enthusiastic responses from the LLM are really going to do a number on people's egos.

[ulbu]

people already go psychotic on LLMs lol

[aeve890]

>MAJOR BREAKTHROUGH ACHIEVED

the bootlicking behavior must must be like crack for wannabes. jfc

>I did not write a single line of code but carefully shepherded AI over the course of several days and kept it on the straight and narrow.

>AI: I need to keep track of variables moving across registers. This is too hard, let’s go shopping… Me: Hey, don’t any no shortcuts!

>My work was just directing, shaping, cajoling and reviewing.

How people can say that without the slightest bit of reflection on whether they're right or just spitting BS

[thorn]

Ah. I remember that guy. Joel. He sold his poker server and bragged around HN long time ago. He is too much of PR stunt guy recently. Unfortunately AI does not lead to people being nice in the end. The way people abuse other people using AI is crazy. Kudos to ocaml owners giving him a proper f-off but polite response.

[jeffbee]

I agree that's a funny coincidence. But, what about the change it wanted to commit? It is at least slightly interesting. It is doubly interesting that changing line 638 neither breaks nor fixes any test.

[quirino]

There's a tweet with a Claude screenshot with a bit more context (linked on the PR).

I don't know enough about the project to know if it makes any sense, but the Zig contributor seemed confused (at least about the title).

[joelreymont]

This one https://x.com/joelreymont/status/1990981118783352952?s=20

I made the mistake of poorly documenting that PR.

[wavemode]

Perhaps the offset is always zero anyway in that scenario

But yeah hard to say

[joelreymont]

It was a bug in my own modes to the Zig compiler not in the stock compiler itself.

[joelreymont]

That one was poorly documented and may have been related to an issue in my code.

I would offer this one instead.

https://github.com/joelreymont/zig/pull/1

[noname120]

Can you stop wasting everyone’s time?

[Levitating]

Just reading his blogposts, gross. He not only thinks he is actively contributing, he thinks he deserves credit.

[greazy]

FYI you're replying to the Joel the person being discussed. On HN he uses the username wagerlab

[joelreymont]

Yep, that's me!

I will look into renaming myself, although don't think HN allows this.

[tjalfi]

You might try emailing them, they've renamed other accounts before.

[debugnik]

Even after the public call-outs you keep dropping blatant ads for your blog and AI in general in your PRs; there's no other word for them than ads. This is why I blocked you on the OCaml forum already.

When I was a kid, every year I'd get so obsessed about Christmas toys that the hype would fill my thoughts to the point I'd feel dizzy and throw up. I genuinely think you're going through the adult version of that: your guts might be ok but your mind is so filled with hype that you're losing self-awareness.

[rs186]

I wonder why the maintainers haven't banned this dude yet.

[Eldt]

Stop the catfishes, please