it's at 10 now. note: the article does not say "taste" once

At this point I'd settle if they bothered to read it themselves. There's a lot of stuff posted that feels to me like the author only skimmed it and expects the masses to read it in full.

I remember in the early days of LLMs this was the joke meme. But, now seeing it happen in real life is more than just alarming. It's ridiculous. It's like the opposite of compressing a payload over the wire: We're taking our output, expanding it, transmitting it over the wire, and then compressing it again for input. Why do we do this?

I had a similar realization. My team was discussing whether we should hook our open-source codebases into an AI to generate documentation for other developers, and someone said "why can't they just generate documentation for it themselves with AI"? It's a good point: what value would our AI-generated documentation provide that theirs wouldn't?

It was especially unfortunate because to do its thing, the code required a third party's own personal user credentials including MFA, which is a complete non-starter in server-side code, but apparently the manager's LLM wasn't aware enough to know that.

>AI frees my human brain to think about goals, features, concepts, user experience and "big picture" stuff.

The trigger for the post was about post-AI Show HN, not about about whether vibe-coding is of value to vibe-coders, whatever their coding chops are. For Show HN posts, the sentence I quoted precisely describes the things that would be mind-numbingly boring to Show HN readers.

pre-AI, what was impressive to Show HN readers was that you were able to actually implement all that you describe in that sentence by yourselves and also have some biochemist commenting, "I'm working at a so-and-so research lab and this is exactly what I was looking for!"

Now the biochemist is out there vibe-coding their own solution, and now, there is no way for the HN reader to differentiate your "robust" entry from a completely vibe-code noobie entry, no matter how long you worked on the "important stuff".

Why? because the barrier of entry has been completely obliterated. What we took for granted was that "knowing how to code" was a proxy filter for "thought and worked hard on the problem." And that filter allowed for high-quality posts.

That is why the observation that you know longer can guarentee or have any way of telling quickly that the posters spent some time on the problem is a great observation.

The very value that you gain from vibe-coding is also the very thing that threatens to turn Show HN into a glorified Product Hunt cesspool.

"No one goes there any more, it's too crowded." etc etc

"Gatekeeping" became a trendy term for a while, but in the post-LLM world people are recognizing that "gatekeeping" is not the same as "having a set of standards or rules by which a community abides".

If you have a nice community where anyone can come in and do whatever they want, you no longer have a community, you have a garbage dump. A gate to keep out the people who arrive with bags of garbage is not a bad thing.

We have always relied on superficial cues to tell us about some deeper quality (good faith, willingness to comply with code of conduct, and so on). This is useful and is a necessary shortcut, as if we had to assess everyone and everything from first principles every time things would grind to a halt. Once a cue becomes unviable, the “gate” is not eliminated (except if briefly); the cue is just replaced with something else that is more difficult to circumvent.

I think that brief time after Internet enabled global communication and before LLMs devalued communication signals was pretty cool; now it seems like there’s more and more closed, private or paid communities.

Whoa whoa whoa hold your horses, code has a pretty important property that ordinary prose doesn’t have: it can make real things happen even if no one reads it (it’s executable).

I don’t want to read something that someone didn’t take the time to write. But I’ll gladly use a tool someone had an AI write, as long as it works (which these things increasingly do). Really elegant code is cool to read, but many tools I use daily are closed source, so I have no idea if their code is elegant or not. I only care if it works.

Amen to that. I am currently cc'd on a thread between two third-parties, each hucking LLM generated emails at each other that are getting longer and longer. I don't think either of them are reading or thinking about the responses they are writing at this point.

DJing is an interesting example. Compared with like composition, Beatmatching is "relatively" easy to learn, but was solved with CD turntables that can beatmatch themselves, and yet has nothing to do with the taste you have to develop to be a good DJ.

I feel like I can breeze past the easy, time consuming infrastructure phase of projects, and spend MUCH more time getting to high level interesting problems?

For every person like you who puts in actual thought into the project, and uses these tools as coding assistants, there are ~100 people who offload all of their thinking to the tool.

It's frightening how little collective thought is put into the ramifications of this trend not only on our industry, but on the world at large.

There are plenty of times in which people will prefer the technically inferior or less aesthetically pleasing output because of the story accompanying it. Different people select different intention to value, some select for the intention to create an accurate depiction of a beautiful landscape, some select for the intention to create a blurry smudge of a landscape.

I can appreciate the art piece made my someone who only has access to a pencil and their imagination more than someone who has access to adobe CC and the internet because its not about the output to me its about the intention and the story.

Saying I made this drawing implies that you at least sat down and had the intention to draw the thing. Then revealing that you actually used AI to generate it changes the baseline assumption and forces people to re-evaluate it. So its not "finding a creative result that they value, but they retroactively devaluing it if it’s not created by a process that they consider artistic

Sometimes you do, which is why there’s not only a single type of brush in a studio. You want something very controllable if you’re doing lineart with ink.

Even with digital painting, there’s a lot of fussing with the brush engine. There’s even a market for selling presets.

and that's because people have a weird sort of stylistic cargo-culting that they use to evaluate their writing rather than deciding "does this communicate my ideas efficiently"?

for example, young grad students will always write the most opaque and complicated science papers. from their novice perspective, EVERY paper they read is a little opaque and complicated so they try to emulate that in their writing.

office workers do the same thing. every email from corporate is bland and boring and uses far too many words to say nothing. you want your style to match theirs, so you dump it into an AI machine and you're thrilled that your writing has become just as vapid and verbose as your CEO.

No one in their right mind would use one.

Using the wrong tool for the job results in disaster.

It's like watching a guy bang rocks together to "vibe build" a house. Good luck.

Maybe it will make them output better text, but it doesn’t make them better writers. That’d be like saying (to borrow the analogy from the post) that using an excavator makes you better at lifting weights. It doesn’t. You don’t improve, you don’t get better, it’s only the produced artefact which becomes superficially different.

> If you're going to be doing much writing, you should have your own prompt that can help with your voice and style.

The point of the article is the thinking. Style is something completely orthogonal. It’s irrelevant to the discussion.

LLMs and agents work the same way. They’re power tools. Skill and judgment determine whether you build more, or lose fingers faster.

No one finds AI-assisted prose/code/ideas boring, per se. They find bad prose/code/ideas boring. "AI makes you boring" is this generation's version of complaining about typing or cellular phones. AI is just a tool; it's up to humans how to use it.

If they don't care enough to improve themselves at the task in the first place then why would they improve at all? Osmosis?

If this worked then letting a world renown author write all my letters for me will make me a better writer. Right?

Who cares if you're a "good writer?" Are you "easy to understand" is the real achievement.

> They are original to me, and that feels like an insightful moment, and thats about it.

The insight is that good ideas (whether wholly original or otherwise) are the result of many of these insightful moments over time, and when you bypass those insightful moments and the struggle of "recreating" old ideas, you're losing out on that process.

The LLM helps me gather/scaffold my thoughts, but then I express them in my own voice

For what it's worth, the unifying idea behind both is basically a "hazing ritual", or more neutrally phrased, skin in the game. It takes time and energy to look at things people produce. You should spend time and energy making sure I'm not looking at a pile of shit. Doesn't matter if it's a website or prose.

Obviously some people don't. And that's why the signal to noise ratio is becoming shit very quickly.

Rest of the world: "No, we're gatekeeping because we think the result isn't good."

If someone can cajole their LLM to emit something worthwhile, e.g. Terence Tao's LLM generated proofs, people will be happy to acknowledge it. Most people are incapable of that and no number of protestations of gatekeeping can cover up the unoriginality and poor quality of their LLM results.

AI is almost the exact opposite. It's verbose fluff that's only superficially structured well. It's worse than average

(waiting for someone to reply that I can tell the AI to be concise and meaningful)

This is a point that often results in bad faith arguments from both AI enthusiasts and AI skeptics. Enthusiasts will say "everything is a remix and the most creative works are built on previous works" while skeptics will say "LLMs are stochastic parrots and cannot create anything new by technical definition".

The truth is somewhere in the middle, which unfortunately invokes the Golden Mean Fallacy that makes no one happy.

This behavior has already been happening with Pangram Labs which supposedly does have good AI detection.

I wonder if this is a major differentiator between AI fans and detractors. I dislike and actively avoid anything closed source. I fully agree with the premise of the submission as well.

"One" is the operative word here, supposing this includes only humans and excludes AI agents. When code is executed, it does get read (by the computer). Making that happen is a conscious choice on the part of a human operator.

The same kind of conscious choice can feed writing to an LLM to see what it does in response. That is much the same kind of "execution", just non-deterministic (and, when given any tools beyond standard input and standard output, potentially dangerous in all the same ways, but worse because of the nondeterminism).

For example, I have a few letter generators on my website. The letters are often verified by a lawyer, but the generator could totally be vibe-coded. It's basically an HTML form that fills in the blanks in the template. Other tools are basically "take input, run calculation, show output". If I can plug in a well-tested calculation, AI could easily build the rest of the tool. I have been staunchly against using AI in my line of work, but this is an acceptable use of it.

But isn't this the distinction that language models are collapsing? There are 'prose' prompt collections that certainly make (programmatic) things happen, just as there is significant concern about the effect of LLM-generated prose on social media, influence campaigns, etc.

In the arts the differentiators have always been technical skill, technical inventiveness, original imagination, and taste - the indefinable factor that makes one creative work more resonant than another.

AI automates some of those, often to a better-than-median extent. But so far taste remains elusive. It's the opposite of the "Throw everything in a bucket and fish out some interesting interpolation of it by poking around with some approximate sense of direction until you find something you like" that defines how LLMs work.

The definition of slop is poor taste. By that definition a lot of human work is also slop.

But that also means that in spite of the technical crudity, it's possible to produce interesting AI work if you have taste and a cultivated aesthetic, and aren't just telling the machine "make me something interesting based on this description."

"Gatekeeping" is NOT when you require someone to be willing learn a skill in order to join a community of people with that skill.

And in fact, saying "you are too stupid to learn that on your own, use an AI instead" is kind of gatekeeping on its own, because it implicitly creates a shrinking elite who actually have the knowledge (that is fed to the AI so it can be regurgitated for everyone else), shutting out the majority who are stuck in the "LLM slum".

If it's not worth reading something where the writer didn't take the time to write it, by extension that means nobody read the code.

Which means nobody understands it, beyond the external behaviour they've tested.

I'd have some issues with using such software, at least where reliability matters. Blackbox testing only gets you so far.

But I guess as opposed to other types of writing, developers _do_ read generated code. At least as soon as something goes wrong.

I gotta disagree with you there! Code that isn't read doesn't do anything. Code must be read to be compiled, it must be read to be interpreted, etc.

I think this points to a difference in our understanding of "read" means, perhaps? To expand my pithy "not gonna read if you didn't write" bit: The idea that code stands on its own is a lie. The world changes around code and code must be changed to keep up with the world. Every "program" (is the git I run the same as the git you run?) is a living document that people maintain as need be. So when we extend the "not read / didn't write" it's not using the program (which I guess is like taking the lessons from a book) it's maintaining the program.

So I think it's possible that I could derive benefit from someone else reading an llm's text output (they get an idea) - but what we are trying to talk about is the work of maintaining a text.

It works, sure, but is it worth your time to use? I think a common blind spot for software engineers is understanding how hard it is to get people to use software they aren’t effectively forced to use (through work or in order to gain access to something or ‘network effects’ or whatever).

Most people’s time and attention is precious, their habits are ingrained, and they are fundamentally pretty lazy.

And people that don’t fall into the ‘most people’ I just described, probably won’t want to use software you had an LLM write up when they could have just done it themselves to meet their exact need. UNLESS it’s something very novel that came from a bit of innovation that LLMs are incapable of. But that bit isn’t what we are talking about here, I don’t think.

The real cost isn't the tokens, it's the attention debt. Every CC'd person now has to triage whether any of those paragraphs contain an actual decision or action item. In my experience running multiple products, the signal-to-noise ratio in AI-drafted comms is brutal. The text looks professional, reads smoothly, but says almost nothing.

I've started treating any email over ~4 paragraphs the same way I treat Terms of Service — skim the first sentence of each paragraph and hope nothing important is buried in paragraph seven.

The most recent one I remember commenting on, the poor guy had a project that basically tried to "skip" IaC tools, and his tool basically went nuts in the console (or API, I don't remember) in one account, then exported it all to another account for reasons that didn't make any sense at all. These are already solved problems (in multiple ways) and it seemed like the person just didn't realize terraformer was already an existing, proven tool.

I am not trying to say these things don't allow you to prototype quickly or get tedious, easy stuff out of the way. I'm saying that if you try to solve a problem in a domain that you have no expertise in with these tools and show other experts your work, they may chuckle at what you tried to do because it sometimes does look very silly.

I see a lot of these discussions where a person gets feelings/feels mad about something and suddenly a lot of black and white thinking starts happening. I guess that's just part of being human.

It's a fantastic editor!

Creativity often requires reasoning in unusual ways, and evaluating those ideas requires learning. The first part we can probably get LLMs to do; the latter part we can't (RL is a separate process and not really scalable).

Even without any of that, you can prompt your way into new things. I'm building a camper out of wood, and I've gotten older LLM models to make novel camper designs just by asking it questions and choosing things. You can make other AI models make novel music by prompting it to combine different aspects of music into a new song. Human creativity works that way too. Think of all the failed attempts at new things that humans come up with before a good one actually sticks.

AI is a bicycle, not a motorcycle.

"You're describing the default output, and you're right — it's bad. But that's like judging a programming language by its tutorial examples.

The actual skill is in the prompting, editing, and knowing when to throw the output away entirely. I use LLMs daily for technical writing and the first draft is almost never the final product. It's a starting point I can reshape faster than staring at a blank page.

The real problem isn't that AI can't produce concise, precise writing — it's that most people accept the first completion and hit send. That's a user problem, not a tool problem."

But what I'm replying to, and the vast majority of the AI denial I see, is rooted in a superficial, defensive, almost aesthetic knee jerk rejection of unimportant aspects of human taste and preference.

The WIP features measure breadth and density of these tropes, and each trope has frequency thresholds. Also I don't use AI to identify AI writing to avoid accusatory hallucinations.

I do appreciate the feedback though and will take it into consideration.

What is that? Do you think PhDs have some special way of talking about things?

As much as I'd like to know whether a text was written by a human or not, I'm saddened by the fact that some of these writing patterns have been poisoned by these tools. I enjoy, use, and find many of them to be an elegant way to get a point across. And I refuse to give up the em dash! So if that flags any of my writing—so be it.

I would argue good ideas are not so easy to find. It is harder than it seems to fit the market, and that is why most of apps fail. At the end of the day, everyone is blinded by hubris and ignorance... I do include myself in that.

This is something I like about the LLM future. I get to spend my time with users thinking about their needs and how the product itself could be improved. The AI can write all the CSS and sql queries or whatever to actually implement those features.

If the interesting thing about software is the code itself - like the concepts and so on, then yeah do that yourself. I like working with CRDTs because they’re a fun little puzzle. But most code isn’t like that. Most code just needs to move some text from over here to over there. For code like that, it’s the user experience that’s interesting. I’m happy to offload the grunt work to Claude.

I got Claude to make a test suite the other day for a couple RFCs so I could check for spec compliance. It made a test runner and about 300 tests. And an html frontend to view the test results in a big table. Claude and I wrote 8500 lines of code in a day.

I don’t care how the test runner works, so long as it works. I really just care about the test results. Is it finding real bugs? Well, we went though the 60 or so failing tests. We changed 3 tests, because Claude had misunderstood the rfc. The rest were real bugs.

I’m sure the test runner would be more beautiful if I wrote it by hand. But I don’t care. I’ve written test runners before. They’re not interesting. I’m all for beautiful, artisanal code. I love programming. But sometimes I just want to get a job done. Sometimes the code isn’t for reading. It’s for running.

This is also the case for AI generated projects btw, the backend projects that I’ve been looking at often contains reimplementations of common functionality that already exists elsewhere, such as in-memory LRU caches when they should have just used a library.

But AI does the code. Well... usually.

People call my project creative. Some are actually using it.

I feel many technical things aren't really technical things they are simply a problem where "have a web app" is part of the solution but the real part of the solution is in the content and the interaction design of it, not in how you solved the challenge technically.

Every single time I have vibe coded a project I cared about, letting the AI rip with mild code review and rigorous testing has bit me in the ass, without fail. It doesn't extend it in the taste that I want, things are clearly spiraling out of control, etc. Just satisfying some specs at the time of creation isn't enough. These things evolve, they're a living being.

"the author (pilot?) hasn't generally thought too much about the problem space, and so there isn't really much of a discussion to be had. The cool part about pre-AI show HN is you got to talk to someone who had thought about a problem for way longer than you had. It was a real opportunity to learn something new, to get an entirely different perspective."

We tell stories of Therac 25 but 90% of software out there doesn’t kill people. Annoys people and wastes time yes, but reliability doesn’t matter as much.

E-mail, internet and networking, operations on floating point numbers are only kind of somewhat reliable. No one is saying they will not use email because it might not be delivered.

Sure... to a point. But realistically, the "use an LLM to write it yourself" approach still entails costs, both up-front and on-going, even if the cost may be much less than in the past. There's still reason to use software that's provided "off the shelf", and to some extent there's reason to look at it from a "I don't care how you wrote it, as long as it works" mindset.

came from a bit of innovation that LLMs are incapable of.

I think you're making an overly binary distinction on something that is more of a continuum, vis-a-vis "written by human vs written by LLM". There's a middle ground of "written by human and LLM together". I mean, the people building stuff using something like SpecKit or OpenSpec still spend a lot of time up-front defining the tech stack, requirements, features, guardrails, etc. of their project, and iterating on the generated code. Some probably even still hand tune some of the generated code. So should we reject their projects just because they used an LLM at all, or ?? I don't know. At least for me, that might be a step further than I'd go.

You might be on to something. Maybe its self-selection (as in people who want to engage deeply with a certain topic but lack domain expertise might be more likely to go for "vibecodable" solutions).

It's about being oblivious, I suppose. Not too different to claiming there will be no need to write new fiction when an LLM will write the work you want to read by request.

Boring is suppose to be boring for the sake of learning. If you're bored then you're not learning. Take a look back at your code in a weeks time and see if you still understand what's going on. Top level maybe, but the deep down cog of the engine of the application, doubt so. Not to preach but that's what I've discovered.

Unless you already have the knowledge, then fine. "here's my code make it better" but if it's the 14th time you've written the ring buffer, why are you not using one of the previous thirteen versions? Are you saying that the vibed code is more superior then your own coding?

How then is it different from the Wikipedia page you linked?

And to answer you more directly, generally, in my professional world, I don't use closed source software often for security reasons, and when I do, it's from major players with oodles of more resources and capital expenditure than "some guy with a credit card paid for a gemini subscription."

Believe me I've had to adjust my writing a lot to avoid these tells, even academics I know are second guessing everything they've ever been taught. It's quite sad but I think it will result in a more personable internet as people try to distinguish themselves from the bots.

I feel like I've been around these parts for a while, and that is not my experience of what Show HN was originally about, though I'm sure there was always an undercurrent of status hierarchy and approval-seeking, like you suggest.

As we give more and more autonomy to agents, that % may change. Just yesterday I was looking at hexapods and the first thing it tells you ( with a disclaimer its for competitions only ) that it has a lot of space for weapon install. I had to briefly look at the website to make sure I did not accidentally click on some satirical link.

There might be other professions where people get more hung up on formalities but my partner works in a non-tech field and it's the same way there. She's far more likely to get an email dashed off with a sentence fragment or two than a long formal message. She has learned that short emails are more likely to be read and acted on as well.

I was dabbling in consulting infrastructure for a bit, often prospects would come to me with stuff like this "well I'll just have AI do it" and my response has been "ok, do that, but do keep me in mind if that becomes very difficult a year or two down the road." I haven't yet followed up with any of them to see how they are doing, but some of the ideas I heard were just absolute insanity to me.

Exactly this. Finding that annoying bug that took 15 browser tabs and digging deep into some library you're using, digging into where your code is not performant, looking for alternative algorithms or data structures to do something, this is where learning and experience happen. This is why you don't hire a new grad for a senior role, they have not had time to bang their heads on enough problems.

You get no sense of how or why when using AI to crank something out for you. Your boss doesn't care about either, he cares about shipping and profits, which is the true goal of AI. You are an increasingly unimportant cog in that process.

It's okay not to memorize everything involved in a software project. Sometimes what you want to learn or experiment with is elsewhere, and so you use the AI to handle the parts you're less interested in learning at a deep and intimate level. That's okay. This mentality that you absolutely have to work through manually implementing everything, every time, even when it's not related to what you're actually interested in, wanted to do, or your end-goal, just because it "builds character" is understandable, and it can increase your generality, but it's not mandatory.

Additionally, if you're not doing vibe coding, but sort of pair-programming with the AI in something like Zed, where the code is collaboratively edited and it's very code-forward — so it doesn't incentivize you to stay away from the code and ignore it, the way agents like Claude Code do — you can still learn a ton about the deep technical processes of your codebase, and how to implement algorithms, because you can look at what the agent is doing and go:

"Oh, it's having to use a very confusing architecture here to get around this limitation of my architecture elsewhere; it isn't going to understand that later, let alone me. Guess that architectural decision was bad."

"Oh, shit, we used this over complicated architecture/violated local reasoning/referential transparency/modularity/deep-narrow modules/single-concern principles, and now we can't make changes effectively, and I'm confused. I shouldn't do that in the future."

"Hmm, this algorithm is too slow for this use-case, even though it's theoretically better, let's try another one."

"After profiling the program, it's too slow here, here, and here — it looks like we should've added caching here, avoided doing that work at all there, and used a better algorithm there."

"Having described this code and seeing it written out, I see it's overcomplicated/not DRY enough, and thus difficult to modify/read, let's simplify/factor out."

"Interesting, I thought the technologies I chose would be able to do XYZ, but actually it turns out they're not as good at that as I thought / have other drawbacks / didn't pan out long term, and it's causing the AI to write reams of code to compensate, which is coming back to bite me in the ass, I now understand the tradeoffs of these technologies better."

Or even just things like

"Oh! I didn't know this language/framework/library could do that! Although I may not remember the precise syntax, that's a useful thing I'll file away for later."

"Oh, so that's what that looks like / that's how you do it. Got it. I'll look that up and read more about it, and save the bookmark."

> Unless you already have the knowledge, then fine. "here's my code make it better" but if it's the 14th time you've written the ring buffer, why are you not using one of the previous thirteen versions? Are you saying that the vibed code is more superior then your own coding?

There are a lot of reasons one might not be able to, or want to, use existing dependencies.

I assume you use JavaScript? TypeScript or Go perhaps?

Pfft, amateur. I only code in Assembly. Must be boring for you using such a high-level language. How do you learn anything? I bet you don't even know what the cog of the engine is doing.

Absolutely, but I’d categorize that ‘bit’ as the innovation from the human. I guess it’s usually just ongoing validation that the software is headed down a path of usefulness which is hard to specify up-front and by definition something only the user (or a very good proxy) can do (and even they are usually bad at it).

I made it because at that point in my career I simply didn't know that ansible existed, or cloud solutions that were very cheap to do the same thing. I spent a crazy amount of effort doing something that ansible probably could have done for me in an afternoon. That's what sometimes these projects feel like to me. It's kind of like a solution looking for a problem a lot of the time.

I just scanned through the front page of the show HN page and quickly eyeballed several of these type of things.

I applaud your optimism, but I think the internet is a lost cause. Humans who value communicating with other humans will need to retreat into niche communities with zero tolerance for bots. Filtering out bot content will likely continue to be impossible, but we'll eventually settle on a good way to determine if someone is human. I just hope we won't have to give up our privacy and anonymity for it.

Agreed.

> I made it because at that point in my career I simply didn't know that ansible existed

Channels Mark Twain. "Sorry for such a long letter, i didn't have the time to make it shorter."

I write software when the scripts are no longer suitable.

What is your hoped for outcome here man? To come off like enough of a jerk or obtuse enough that people just abandon the thread and you can declare victory?