Hacker News Reader

Comment by chicken-stew | Hacker News Reader

Not so long ago I read something on HN which resonated with me: we’re moving into the era that is comparable to car mechanics enthusiasts. You have a previous generation of cars where people just enjoy working on with hand tools, and there’s modern cars where people like to tune with software patches.

From the very start of llms I’ve had nothing but bad experiences with code that was generated for me. Either it’s buggy, it works but I end up losing an evening on some obscure bug, or it’s full of red flags.

My latest hobby project is just in a text editor with markup and that’s it. I’m also done with the augmented assistance in the IDE. I google things I forgot. I constantly read these amazing stories of people vibe-coding some firmware/driver that just works, and honestly I’m starting to question whether I’m reading the posts of some promotional bot.

View on HN

For me it's just making stuff for my family and kids and seeing if it sticks. LLM's have given me a way to easily get something off the ground whereas it used to be a few months now is a few evenings.

I've made a wealth of cute things for me and my kids to do (shameless self promotion) https://drawesome.art - multiple people drawing/painting/coloring on the same canvas my kids love this. The architecture took forever for me to get write. Basically server side authoritative state of pixels and blending between two points. How to realistically manage and store it. https://www.catchmemeow.com - random where's waldo-esque game where you find a cat or dog wearing a silly hat I made this for my daughter and me to help the shelter she wants to volunteer at. https://www.cluestep.com - this one is weird I've actually tried to productize it but i made this because I had the HARDEST time helping my oldest with her calculus/algebra II homework. Basically I constantly made her cry. I just wish I could make it not look so "Made with AI"

I think LLM's analogy are prefab with nailguns did for carpentry. It's very complicated to prompt it (at least for me) to try and make it do something novel and net new. It always falls back on node/typescript/fastAPI/python. It's not terrible at using more obscure languages but I have the hardest time describing what I want or thinking without starting something more by hand and then letting it take the wheel and seeing where it drives me.

I should mention too at work I'm practically forced, measured even on the amount of AI I use. The expectation feels like use AI for any and everything. So i'm more worried one day I'll be laid off once I'm found to hit the 'enter' key much anymore. Anyways for me I'd say embrace it, use it and you'll find your output is much higher with the caveat that when a bug occurs it's incredibly hard sometimes to fix if it's with LLM generated code.

I'm hardly AI-pilled but these are lovely parent/child apps. Kudos

Lowering the barrier to proof of concepts for personal projects is a beautiful thing. I don’t love AI but I’m 100% playing the cards I’m dealt.

It just dawned on me that LLM use is a lot more measurable then "the rockstar engineer spent a week bouncing their favorite nerf ball off the wall all week and then the solution came to them" and the business side always loves to measure productivity metrics etc. As someone who's worked for mom and pop and big corporate the stats spook me, or if they're available to the worker I find them more distracting and encourage gaming the stats then being productive or humane to your co-workers.

> "the rockstar engineer spent a week bouncing their favorite nerf ball off the wall all week and then the solution came to them"

I wouldn’t call myself a rockstar but I did get a job after a summer internship which I mostly spent touring with the Grateful Dead. At the end of the summer I was in the office and the head of the research division said he hadn’t seen me much(!). I said I’d been thinking and had some ideas on how the group’s research plan was flawed and how to do it better. I then filled his whiteboard with off the cuff ideas.

I spent the next several years with a team implementing the whiteboard.

I like to think things work better these days, AI or no AI

I've written about this at length on my blog[1], but it's sorta impossible to measure the work that goes on in ourbrains. So we just say "anything we can't measure doesn't matter".

[1] https://blog.nilesh.io/post/llms-and-jobs

Idk, I could see some of my stats (like Claude usage) and I tried to avoid spending an unjustified amount of money.

There are times when I wonder if a model is just spinning its wheels to burn tokens, and how hard would it be to write that loop from the API provider side... Lex Luther shower thoughts!

You're on to something

> this one is weird I've actually tried to productize it but i made this because I had the HARDEST time helping my oldest with her calculus/algebra II homework.

I feel somewhat qualified to opine on how humans learn because I've successfully taught my 2x boys[1].

Teaching is difficult. It requires a lot of patience and understanding, and a working mental model of the human brain.

If you have tried, and failed, teaching calculus to your kid, then maybe you don't have a good model of how human brains work.

TBH, it doesn't even have to be correct, it just needs to work; celestial navigation had the wrong model for centuries, but ships still got to where they were going (mostly).

I'm finding that developers claiming to have learned something have often very superficially learned it. If they were tested on what they learned, they'd fail, just like the HS trig students who read the textbook and did no problems fail when they hit the exam.

--------------------------

[1] My youngest, at grade-1, reads at a grade 3 level and can do things like multiple 27 by 4 in his head. That is a result of a structured and scheduled 10m lesson per day since he was 3.5 My oldest was getting 5% in trig/geometry in grade 11 (long story behind that one), and became solidly mid-70% after my tuition.

Yeah I have 4 kids. I'm come to the realization all brains are different. My daughter struggles with anxiety and feeling terrible if she doesn't immediately 'get it' especially in a social setting. She does great with online tools where she can feel like no one is judging her.

My son doesn't get upset at all with me he'd just play his mom to get the help to the right answer and didn't have the 'want' to understand a problem he just wanted to move along.

I have another daughter who is autistic. I'm still trying everything I can possibly do to get her past reading at a 2nd grade level. That's another problem I'm trying/struggling to solve.

I feel the same way. Mostly with tools for people though. How do you bring people closer in the age of social media and ai. I keep seeing things like https://www.wishlst.com and https://journalee.com and trying to brainstorm things that would be helpful to humans regardless of tech.

  > You have a previous generation of cars where people just enjoy working on with hand tools, and there’s modern cars where people like to tune with software patches.

I think this is a prettier picture than reality suggests, it's overly simplified and a statement people like because it's a variant of "everything is the same"

I think the analogy with cars is apt, but we should also like at what happened with cars. People are being locked out of fixing them, even for simple jobs that would usually be task that introduces someone to the hobby or career. Fixes like changing your oil or brakes. Remember that the Ford CEO said that cars are "too dangerous" to fix. Probably connected with the other thing they talk about... a $100 bn repair industry that they want a bigger piece of. There's an easy way to do that, John Deer and Apple have clearly shown that it's not hard, even when we get the right to repair. Is it any wonder far fewer people work on their cars these days? Ask the people with electric vehicles... there are people out there, just as there always will be, not the numbers do matter

Weirdly, I think LLMs would be much more useful if everything was open source. No need to hack your microwave to give it a firmware update to the door doesn't lock, just patch. More training for the LLMs and turns what would be a specialized task into one anyone can do if they point the LLM at the code. Plus, people submit patches to the source code, getting companies free work. There is a way everyone can win in this. Sometimes (often) being greedy prevents you from getting a bigger fill

It might be that you're just working on projects where LLMs don't provide much benefit.

The second thing is that using LLMs in a useful way is a skill in itself.

> The second thing is that using LLMs in a useful way is a skill in itself.

I have to disagree here. People who are really into agentic coding like to say/believe this, but come on, it's not rocket science! You literally ask the agent to do something and it does it, that's it. As long as you use a recent model it will do it really well regardless of how fancy your prompt is. There's no moat here, anyone can do it. Your job will be gone and so will mine.

It's not hard to use a drill or a hammer but learning how to use them as tools to make things was always the real challenge. Using LLMs in a useful way I would argue is a much more important skill than programming as was learning how to make programming useful was always the more important skill, but it depends on if you view programming as a means to an end or the end in and of itself.

Using LLMs in a useful way I would argue is a much more important skill than programming as was

The trouble with this argument is that fundamentally this whole field has no sound theoretical foundation. It has no deterministic rules you can learn that will reliably get good results.

That also means there is no way to prove that what works well today will continue to work well tomorrow. Nor can you be sure that something you've tried and ruled out in the past because it didn't get satisfactory results won't be the new SOP next week. Any experience you gain in the field might be obsolete within hours.

An important consequence of this is that you can't train someone else to use LLM-based tools effectively either. At best you can train them to use specific tools and models that exist today effectively and hope the knowledge remains useful for a significant period of time. Even that assumes you have the knowledge yourself and aren't relying on something that worked well yesterday but could already be out of date after something new was released last night.

This looks like a fundamental problem that can never be fully solved as long as LLMs of the type that are popular today are behind the tools. People forget in all the hype and rapid change that the whole idea of agentic AI tools in mainstream development is barely a year old. We have no idea yet what the long-term consequences of so many people and organisations in the industry all but abandoning traditional programming skills in favour of AI agents will be. There already seems to be a lot of anecdotal evidence about programming skills atrophying, developers burning out, and the quality of the finished product dropping but it's probably too soon to have serious data to analyse or to draw any big conclusions about what is a good way forward from here.

Then on the other hand you have people who have doubled or tripled their real-world business by using these tools… And the folks who have developed apps that benefit their child’s development, etc.

You don’t have to look too far to see where we are going. But you do need to have your eyes open.

You don’t have to look too far to see where we are going. But you do need to have your eyes open.

As I said - we have no idea yet what the long-term consequences of the recent rapid shift towards AI and agents will be. I prefer to make decisions based on evidence, not hearsay or wishful thinking.

I have been around long enough to see a lot of hype cycles in programming. They were all going to be revolutionary. They were all going to change the nature of programming forever. They were all going to result in dramatic increases to productivity (or other similar claims).

Many of those phases did turn out to have some good - sometimes excellent - ideas. A lot of those have since been widely adopted in the industry. They also had plenty of ideas that didn't really work out. In the end we have improved some areas to a useful degree but none of those phases resulted in orders of magnitude increases in useful productivity or quality.

So far LLMs and agents are looking like a familiar story. They're proving to be useful tools for some types of work. They're now quite good at producing code to the standard of an average developer doing a well understood task - and in many cases that is all you need! But the jury is still out on whether they'll ever be able to replace good programmers or build genuinely innovative products where there aren't lots of examples of good existing implementations to train the models. Those are the parts of the industry where I do most of my own work and I hear much more scepticism about these tools from those developers than in the average online discussion.

> Using LLMs in a useful way I would argue is a much more important skill than programming as was learning how to make programming useful was always the more important skill, but it depends on if you view programming as a means to an end or the end in and of itself.

No, it's not. As evidence of intelligence atrophy, I offer up this word salad that barely makes any sense as evidence: someone barely able to write a coherent thought can apparently produce working software.

>I have to disagree here. People who are really into agentic coding like to say/believe this, but come on, it's not rocket science! You literally ask the agent to do something and it does it, that's it. As long as you use a recent model it will do it really well regardless of how fancy your prompt is. There's no moat here, anyone can do it. Your job will be gone and so will mine.

As someone who's used more and more software made this way—either for hobby stuff or because I've had to for work—I just don't get it. You don't always get exactly what you ask for, and you often ask for something that is really sub-optimal in many ways.

The problem with waterfall wasn't ever solely "it takes a long time to write the code."

>You literally ask the agent to do something and it does it

It's not that simple, you have to remember at the end of the day these things are just doing next token prediction. If you don't give it the proper tokens to attend to, then your outputs won't be satisfactory.

You can get stellar outputs from LLMs, but it really is a function of how well you manage your input tokens.

that is to say, you have to be a good dev first and foremost, and no matter how much harnessing you do there is gonna be a limit to the quality of output by the user (otherwise why not just use a no-code tool instead)

(the ironic thing is tho, people starting out with just llms will probably never progress to 'good dev' in the first place...)

This is the one thing I'm curious about, what's the next generation of devs going to look like? Just a bunch of people who treat the codebase as a black box, pump out slop, and call it a day when the feature works?

I imagine there will be value in being able to understand every line of code for a while, because many (most?) things don't seem amenable to long term vibe coding. But how will that knowledge be obtained? Maybe there will be a much smaller pool of devs with the patience to actual learn bespoke software development.

Or, AI will actually get good enough that vibe coding is better in every case than mindful development. I guess the main factor is how long it takes for that to happen.

  > Just a bunch of people who treat the codebase as a black box, pump out slop, and call it a day when the feature works?

my guess is yes, because that is what i am seeing today in front of me with new grads

  > Or, AI will actually get good enough that vibe coding is better in every case than mindful development. I guess the main factor is how long it takes for that to happen.

i think there is a limit to this; you could train an llm to never dereference a null pointer or something like that, and you could probably minimize other security related stuff, but in the end its a token predictor, so you stuck with probabilities... i wouldn't vibe anything that had any security/legal/user related things but thats just me.

Most of us effectively function as next-token predictors, in my experience. Some of us are trained not to dereference null pointers, but there's always a non-zero probability that we do so anyway. To reduce that risk I use the same tools an LLM would, i.e. code review, automated testing and verification, etc. There isn't obviously much difference between humans and LLMs in that sense.

Like you I would not vibe code something important, but I think that's primarily because I want to have a deep understanding of a system before I accept responsibility for it. An LLM would still be useful during implementation and likely to increase the quality of the result.

It seems very likely that AIs will become capable enough that they can take complete responsibility for development, deployment and maintenance of the systems we're used to building. Further, it seems likely they'll end up much more capable than humans have been. I think that's an eventuality we need to be prepared for.

A well-utilised AI tool can probably already build a much safer and more robust security/legal/user related system than you or I can, in any given time budget. There's a question around what happens as humans are required less for the "well utilised" part, and who is responsible when humans are not really involved anymore.

If you have watched a normie use AI and thought See?! It's not rocket science! We are all the same. then buddy, you are a normie. They are very very bad at it. (If you have not watched an average someone use AI, then I highly recommend it, if for no other reason than to understand your placement on the curve.)

> Your job will be gone and so will mine.

I mean, that's probably true, but the person you are replying to is talking about the present.

The skill is in recognizing where they go wrong, which is indistinguishable from the skill of being able to do the job without the LLM.

> I have to disagree here. People who are really into agentic coding like to say/believe this, but come on, it's not rocket science! You literally ask the agent to do something and it does it, that's it. As long as you use a recent model it will do it really well regardless of how fancy your prompt is. There's no moat here, anyone can do it. Your job will be gone and so will mine.

Sure, the skill of writing beautiful for loops and recursive binary search tree is basically worthless, but software is written for a purpose. That purpose is not gone. People will keep wanting POS system, accounting system, lawyers, manufacture stuff, personal assistant, etc etc

> You literally ask the agent to do something and it does it, that's it.

When was the last time you built something meaningful with it? I admit, LLMs do help and save tons of time and effort, yet building noteworthy software still remains a difficult task, with or without LLMs.

Funny, but your comment reminded me of my wife. Circa 2009, she was walking behind my back, she stopped to watch me work without me noticing. I was using Visual Studio (sluggish, temperamental mammoth, not its smaller cousin). She stared at my screen for a few minutes and suddenly exclaimed: "You're not even working, this thing is telling you what to do. I could probably do this shit too..." By "this thing" she obviously meant the intellisense completion that was giving me hints whenever I typed.

It very much is a skill. You wont get good results by asking it to "go do X" and walking away from the desk. You have to know how to use it as a tool to: 1) research the task 2) develop a plan the agent can follow to complete the task 3) check the plan for completeness/issues

After a (potentially long) session of investigation and planning _then_ you let it code. Even just being aware of this process is part of the skill involved (then you have to actually get used to doing it effectively).

> It very much is a skill. You wont get good results by asking it to "go do X" and walking away from the desk.

You soon will. If not next year, then maybe the year after, or the year after that.

Whatever value you are providing will evaporate in a few years.

Are you planning on stopping eating in 3 years?

Ask a human to "build twitter" and walk away and tell me how it goes. Vague requests will never work, no matter how intelligent the person being asked is, because they don't have enough information to give you what you want.

"No no no, when I said build twitter I meant without the character limit!"

What's your point? I can't do anything about that. All I can do is deal with the situation as it is right now and wait to see what happens.

Dealing with the situation would mean actually acting in your self interest instead of deferring to a cookie-cutter narrative to justify yourself as an essential meat proxy. Maybe stop feeding your data into the job destroying machine while paying for the privilege? Maybe come up with a better story for management than prompt wispersing, like at least writing your integration code yourself...

> You wont get good results by asking it to "go do X" and walking away from the desk

but you will!! you absolutely will!

> You literally ask the agent to do something and it does it, that's it

Exactly, you have to know what to ask, especially on large codebase. I still have to baby sit Astra and opus 5.5 into reusing existing structures/functions, not reinventing the wheels, telling them "yes this might be a legit race condition, but you added 5000 LOC and I'm pretty sure it will never ever be a problem in production given it would require 5 distinct catastrophic failures at once to trigger it, if it ever happens log it in sentry and we'll have a look by then".

Some people are really really bad and won't even think about asking stuff like "run a memory profile and see if we can decrease our footprint", "benchmark the top 3 options for this bug and keep the one that uses the less cpu but still respond under 150ms".

You still need to have a broad understanding of what a computer is and how it works, both are skills, which a lot of dev didn't have anyways, but now they can shit out a looooot more code

My wording was bad. I've been using latest models recently and I can pretty much say build xyz feature or fix xyz bug and it will do it really well. It will ask me any clarifying questions. It will highlight things I haven't thought about. It will make suggestions on how to test and verify. And all of these things it will highlight the recommended choice I should choose.

So I stand by my assertion that anyone can do this now, it's definitely not a hard to learn/master skill, and people who say it is are naively believing there is a moat around their career.

So people who choose to not use it are not losing out on building any skills- in fact if they are focusing on hand coding and learning and keeping their brain sharp they, in my opinion, will be in a much stronger place than the AI enthusiasts whose brains are rotting.

You’re so profoundly incorrect that it’s obviously not worth the time and effort to even begin to try to realign you with reality on a site like this. I don’t know how to educate others on this topic. It’s a serious problem and the gap is continually broadening.

> The second thing is that using LLMs in a useful way is a skill in itself.

Not really. I don't even bother writing complete sentences as my prompts anymore. Once the LLM has enough context on my current issue, even incomplete sentences or phrases result in the same outcome quality in LLM's responses.

I just recently replaced LSP with ctags. I am loving it. I plan to write a post about it but I’m leaning heavily into grep and search tools to navigate code. I do miss some autocomplete and auto import features but other than that I can actually navigate through code better with ctags instead of go to definition.

For linting style errors I’m using entr in a separate terminal which is also awesome.

Please write this up! I’ve had similar ideas but haven’t tried going back to ctags yet. I’d love to read about your experience.

I'm going to get left behind as a developer I guess because I still haven't seen output in my niche that's done well enough to just let it go and develop entire projects. I essentially just still talk to it like it's a rockstar dev and it gives me good suggestions and technical talking points on why one paradigm might be better than another.

Hyped by those stories, I connected an agent to my old Chinese xiaomi phone and asked to install lineageOS on it.

After several days, it managed to brick it and unbrick it. The boot loader is still locked.

I just use plain old vscode with no AI integrations at all.

> I constantly read these amazing stories of people vibe-coding some firmware/driver that just works, and honestly I’m starting to question whether I’m reading the posts of some promotional bot.

No one uses assistance in the IDE for project like these. They're all using Claude Code (or similar harnesses) and lots of sub-agents. People who haven't seriously tried out SOTA models + their proprietary harnesses just don't know what they're missing out.

> constantly read these amazing stories of people vibe-coding some firmware/driver that just works, and honestly I’m starting to question whether I’m reading the posts of some promotional bot.

I don’t question those posts, but these days, I can code something with vi (nvi) and be ok with nothing other the small motion helpers. These days, whenever I see a project with more than a dozen files, my gut tells me there’s something in there that ought to be a library.

> I google things I forgot.

The problem here is that search engines are trash now. The AI companies are going to have to start running their own (if they haven't already) for their models to use that haven't been nerfed. That might be the next moat against open/local LLMs by cloud LLMs, so the search engines are probably going to get even worse.

> I constantly read these amazing stories of people vibe-coding some firmware/driver that just works, and honestly I’m starting to question whether I’m reading the posts of some promotional bot.

There is a lot of bullshit out there for sure and I'm still not convinced llms are a net positive for society (even ignoring energy use and hardware inflation) but it's clear that it works in a lot of areas, for example when it comes to reverse engineering it pretty much is black magic, the progress made in ps5 and other consoles emulation is out of this world

Was happy to see someone forked Zed into "Gram". It was very disappointing to see every code editor go the route of being more than a code editor or even full IDE. I can literally just invoke a separate AI app to do code if I want that. I don't want to have to be actively conscious of every keystroke being shipped up to a backend to be processed. Sometimes I just need a dumb scratchpad to plop a password onto with other shared notes while I'm mid-task in an unsaved tab.

Forgoing the ide is quite a choice

> I constantly read these amazing stories of people vibe-coding some firmware/driver that just works, and honestly I’m starting to question whether I’m reading the posts of some promotional bot.

No, it just works after a few iterations.

And you'd know this if you would use AI instead of working in a text editor with markup.

To each their own, and I understand if working by hand is more fun for you.

But in my opinion you guys don't get to make strong claims about the quality of works that make heavy use of AI.

You don't have the experience, and I bet neither does the guy who wrote the article about it being essential to handwrite all the code after planning. The arguments about the environment and sustainability do not seem relevant.

Yeah I don’t understand. This is late 2026. Opus 5.5 can literally build anything I throw at it. We’re still seeing people who have not been satisfied with _any_ AI generated code?

But what are you throwing at it?

I've thrown a 20 year old .net 4.8 codebase that's a mixture of proprietary sdks and legacy code and LLMs choke. Building some greenfield pieces for the same application works great so far though.

I've also made some plugins for some hobby software with very limited viewing of source code that I'm very happy with.

people get AI psychosis because they see tailwind-infused gradient dashboards and think every codebase has much smaller problems than this, so AI can build everything else

we have a 6 year old codebase with pretty much everything proprietary and there's no AI that can keep up without having an entire team to create an in-house RAG and throw thousands per day in tokens and spend expensive developer time on reviewing and asking for stupid changes that don't fit in

these arguments would make much more sense if tokens were practically free and LLMs could run in pc graphics cards, but when the model requires a data center that can be seen in space just to give you 100 tokens per second then you're basically a car in 1900 when all the roads were dirt roads for horses

Sure, they can build “anything”, but the quality is not good. I have recently played a few vibecoded games I found on F-Droid and the implementation sucked.

LLMs could write perfect code in 2 seconds every time and you'll still be on the hook for figuring out what makes something fun to play and nice to use. In fact maybe even more on the hook since you wouldn't get to procrastinate by working so much on impl details.

I believe there are many out there who are content to publish a barely working app or game.

For apps, I know that the bugs and UX problems can be ironed out. Not in two hours and by the AI alone, but by days (or months, depending on scope) of manual QA testing.

Whether this is possible for games I cannot say as I have very little and only outdated experience with developing games with AI.

Reply on news.ycombinator.com