Doesn't sound like AI is necessarily the root the problem. If he's not the person to test, iterate, and deliver quality with AI, he probably wouldn't be without it. You just figured it out sooner.
Or, if you have reason to believe he's got what it takes - treat it like an leaving opportunity, give him room to make mistakes. This is, ime, what a rookie AI user mistake looks like.
It is easy to get carried away by the momentum. Many have fallen into that trap, not all are irredeemably bad software engineers.
> If he's not the person to test, iterate, and deliver quality with AI, he probably wouldn't be without it
I see the take quite often on HN, and I have a major gripe with it. Yes, people were lazy and we had these problems before LLMs hit the scene, but clearly there has been a sea-change in both the frequency and degree of this behaviour since.
A problem can both exist independent of a catalyst and be exacerbated by it. Arguments that try to window-dress this away are not particularly productive.
It's a sea change because a lot of this code looks plausible and seems to work to someone inexperienced at assuring quality, and it will take some time before the people emboldened by that learn.
I just yesterday had an agent produce a prototype that appeared to work and that looked plausibly well engineered, except I knew from experience that what it had done should be doable with a tenth of the codebase.
Someone without experience might've just accepted that prototype. It took one prompt to get it turned into a reasonable MVP, but you'd need to know what to ask, and you'd need to know you needed to ask.
I agree with your sentiment. Simply saying 'AI' is not the reason, clouds what is actually happening. To illustrate I suggest playing the word swapping game. Swap AI in the sentence "AI isn't the problem" with, guns, drugs, gambling, or dare I say social media!
Human behaviour is complex. Patterns of behaviour that occur in general (there are always outliers of course) dictate so much in society, culture, politics and inter-human relations. Changing these behavioural patterns will affect meaningful real world change.
Social media has affected change in human behaviour. I suspect AI will too. We are starting to see the first shoots of the changes AI will bring.
Human behaviour on the average will choose convenience over effort. It is very difficult to buck this trend societally. To be clear, AI provides an immense amount of convenience.
I think the argument is quite valid. OP writes "business partner". It begs the question - would their business partner get on board if they didn't have a AI subscription by their side? How about instead of 5K PR, OP got the same half-baked buggy crap or absolutely nothing instead? We don't know the business partner, but we are ready to throw the problem at AI agents right away. Well, I find this not particularly productive.
> but we are ready to throw the problem at AI agents right away
I'm not sure I understand this correctly, but I assume it means blaming the agents? If so, that's not my point - I'm describing how many people are using them to be lazy, which shouldn't be particularly controversial.
To your other question: if OP had gotten half the size PR to go through, that sounds more manageable to me? If they had gotten nothing, that could be a useful signal about the capability of their business partner, worth knowing now rather than later. Could one imagine a situation where AI enables a thoughtful business partner to provide more value? Absolutely. Should this be the default assumption? I'd argue that it absolutely shouldn't.
we started this project before was mainstream (we got sidetracked by other work) and him having less experience but being a very good programmer (I went with him to college and worked together for years), he would listen to my input, fix, and adjust. I have decades of experience.
Really sorry to hear so. Your story reinforces a fear I have that humanity is being squeezed out of all manner of things (pre AI too, but even more so since).
> With AI the solution to it is also trivial since an extra three seconds spent writing a prompt could have had the agent run and verify it.
This just recurses the original problem by an additional step: assuming someone is even bothered to write the prompt, they are just as unlikely as before to check what it spits out.
AI may be a productivity amplifier for some, but I don't think it's far-fetched to describe it as a laziness magnifier for the majority.
AI is not the problem here, but the difference in deep beliefs of the company founders.
When I had partners in my business I studied them first deeply like I was going to marry them because I WAS going to marry them!
You should agree on your values or else the company will go nowhere. One founder will raw one way and the other the opposite making the boat to stay in the same place.
I have a software company and the opposite problem: I am obsessive about testing and meta programming. We write way more testing code than code, we automate everything with meta programming.
It was clear several times that we tested and automated way more than necessary. We automated things that only used once for example(although we used later multiple times on other projects).
So we had not the usual problems than companies have like bloat or security vulnerabilities but we had the problem of having software that was was super expensive because we did double or triple the code than the competition with tests that developers hated(they are boring to write). Our software took also way more time to market.
Companies love to buy the cheapest they can today and tomorrow we will see, so that put us in a difficult position. It was specially hard at first, but then this principles gave us an edge(because most companies would compromise early and not reap the benefits later).
Had I a business partner with a different mindset and beliefs, it would have been impossible to me not to change course of the boat in hard times.
BTW we use LLMs with great success. It let's us test even more spending way less than before.
Thanks for your input. I find testing (especially FE, where a single change can ripple through the whole chain) is actually a great way to use LLMs
It's sad because I've known this guy for 20 years and since the beginning he had a tendency towards "work smarter, not harder", a good motto at the time you're 20 but I often saw it crippling him in the long run, while having short-term gains.
Long-term, funny enough, he said he wanted back into this project because he was very unsatisfied with "easy money" (we both got into SEO at the right time and made a ton of easy money) and wants something stimulating for his brain.
In the meantime I had kids, worked my ass off to get the health and fitness level I want (he also wants it... alas nothing changes).
I'm kind of quite committed to this and we're maybe 30% away from finishing, but I almost debate sometimes if I should just do the FE vibe-coded and let him see that it won't be as good as I can manually do it (FE is hard in the end). Just get it out of the door, keep the customers that we have and let it kind of die down.
OP and long time lurker here, I'm desperate and looking for honest-to-god advice on what to do.
I am one of those neurodivergent people who loves programming because it gives me the right stimulation.
I also enjoy drawing, gardening, and exercise.
I don't own a car because I love walking, and such.
So I'm wondering if it's time to call it quits with programming (for a living).
Luckily I have some savings and I have other interests from years of doing random jobs and businesses, and as much as I love programming, especially with others, I think very few of us still want to program.
I'm like the Last Samurai of programming while everyone is using AI-assisted gattling guns.
The differentiator right now is taste. Most people have none (like the business partner in the OP), and some people have plenty (like you seem to have). These two sets of people will produce wildly different results with the same tools.
Two things you can do regarding programming here.
1. Be a consultant for a high wage, use AI, produce good results quickly - little to negative intellectual joy here. You still have the edge here due to good taste.
2. Produce 'artisanal' code in your spare time. Work 3 months for pay, then 9 months on whatever projects scratch your itch.
I still want to program and I don't think it's time to let go. Hit me up (email in bio) if you wanna chat, you sound like a really interesting bloke.
1. is a bit contradictory or at the least a "devils in the details" wishful fantasy which hasn't been showcased convincingly in the wild yet (by "the wild", I mean software eng shops and corps). It's assumed the devs who have taste were able to implement their taste fine pre-AI, sure, but the whole gotcha in 1. is that the ones with good taste currently are not confident in implementing their good taste while simultaneously "producing good results quickly using AI".
I also really enjoy programming, and relate to your comment below about sometimes waking up in the middle of the night because your brain figured out how to code a hard problem.
Sorry to hear you're having a rough time with work. I do think AI availability has made otherwise smart people put out lazy work. But I also think that there are more people who still "code by hand" than you think.
You might find some joy by going to certain events where people talk about different items coding-wise. I enjoyed the talks for this conference recently https://www.youtube.com/@softwareshouldwork (particularly Richard Hipp's, Delaney Gillilan's and Carson Gross' talks), maybe you can find similar spaces. The benefit has been Youtube's algorithm recommending me better software talks now too.
> who loves programming because it gives me the right stimulation
A small (and a bit tangential) personal story: I thought I am one of those people who care about the result, more then about the process.
But, I've recently played a game called "The Farmer Was Replaced" that someone mentioned in the comments in one of the posts here, on HN.
The game is basically a farming sim, except you farm by programming a drone using a Python-like language.
Of course, I didn't use AI, because where is the fan in this? In "normal life" (work, personal projects), it is very hard for me to convince myself to do anything manually, because generating code is more efficient. But in a game, using AI assistance would be like using aimbot in a shooter or a cheatcode in some other game. Yeah, it's much more efficient, but what is the point? There is no challenge and no fun that comes from overcoming the challenge.
So, anyway, I was writing code by hand for the first time in I-don't-even-remember-how-many years. And it was so much fun. I don't remember when was the last time I felt so engaged, so immersed, so "in the flow".
So, yeah. After all, maybe I'm not a ruthless pragmatic who only cares about the final result and doesn't gives a shit, how did he get there...
I booted my MiSTer into TI-99/4A mode a week or so ago and started writing little programs in TI BASIC. TI BASIC sucked. It was slow, due to being written in a form of bytecode which had to be interpreted (and it would in turn interpret your program). And you couldn't do sprites, not without getting the Extended BASIC cart.
But HOLY SHIT THE DOPAMINE RUSH. Just crafting things and making them appear on the screen with a few statements... It was like I was a kid again. I remembered why I liked programming.
It's like... AI is a cheatcode. It's iddqd. Sure, you can cakewalk through a level with it on, but are you really playing the game, then? You certainly haven't gotten better by doing so, and you haven't experienced the involuntary buttcheek clench when you hear a cacodemon hiss from behind you... The whole point of programming is the new things I can do by having gone through the exercise of figuring it out. The ways I've grown, the glimpses I've gotten into the world.
Just find a better place to go. If you have savings, quit this nonsense and start looking for something else while working on yourself.
I just got back to working on a project I like where we don't use any LLMs for code generation because we think it kind of sucks and leads to bad results unless it's for incredibly small contexts (which is usually not worth the prompting).
You're also not as alone as you think in actually liking programming enough to do it by hand even after having used LLMs. There's also the fact that you build more theory up by actually writing the code anyway, so someone who actually wants to know their code base intimitely will always have to write the code.
No it's absolutely not time to let go. You are one of a limited set of people with pre-AI taste and intuition for good software. They aren't making any more of them. And you'll be a super valuable asset anywhere you go.
I mean, if you want to retire go ahead, but don't do it because of AI. AI is just a tool to generate code faster. It doesn't replace the work of actually making design decisions and knowing what will scale and what's better going forward vs what will work now but spiral into unmanageable spaghetti code soon (yes AIs also have a harder time managing messy code than neat code)
I often read here comments along the lines of "there will be plenty of jobs for AI-unfuckers" [1], but I fear the truth may be closer to "sometimes things are broken, are then fixed with a broken solution, and will remain broken till the end of time".
In my last round of job search this year no one asked me questions about good software practices but they did ask me how proficient I was with AI agents. And let's not forget that we are part of an industry whose motto is "move fast and break things". Sure, NASA has a zero-bugs approach to software dev, but they don't have room for all of us.
I'm am certainly an "ai believer" and have been since I saw the tragectory we were on.
It was pretty simple to project the compounding effects and headroom that we had. I go back and forth on this all the time.
There probably won't be much demand for "un-fuckers". Up until recently I thought knowing what to build for a given context would give you quite a bit of runway.
I still think knowing what you should build will be important for the next 12-24 months.. but I have to be honest.. after that.. I don't know.
I don't know how to feel myself. I'm certainly an "ideas" guy as well as a nitty gritty build it guy. I just think my technical side is depreciating very rapidly.
Completely agree. Saying this as someone who loves handcrafting code and has been doing it professionally for decades, it looks like a dead end at this point. LLM generated code might not be pretty, but these damn models are already so good at handling complexity, that they can actually work fine with a byzantine code base (which honestly would also describe most human-derived codebases I have worked with). I am still reeling from this realization and trying to figure out what it means.
SV echo chamber here it seems. No, most businesses globally don't work in idiotic and often amoral mode of 'move fast and break things' (and unlawful on top of that, but who has time for some boring stuff like that when greed fully takes over).
Banks will not work like that for next 10 years at least. The tiny amount of money one could save on those big banking projects by using llms aggressively will bite back immediately when shit just doesn't work, clients leave for good and regulators will fine you to the ground (or remove your banking license). Many other businesses who are not race to the bottom and realize how crucial reliable IT is to their business also won't go that way anytime soon.
World is not ending, there is some transformation and thats about it. Its true there will be much less jobs for devs but its mostly affecting newcomers. Statistical models are not making code better than experienced humans, and won't for some time if ever.
> most businesses globally don't work in idiotic and often amoral mode of 'move fast and break things'
Not sure what business you are in but the non tech industries (corporate IT, in house apps) have not, in general, been associated with quality. Cost cutting, outsourcing to the lowest bidder: yes. Quality: no.
Mostly banking but dipped my feet in many other areas in the past (telco, insurance, municipalities, army etc).
Its true budgets are often the problem, but not much 'move fast and break things' mantra of SV, rather 'move as much as low budget and corresponding quality and bureaucracy allows you to'
This is what’s know as “cope”. People are not being hired on the basis of how well they know how to not use AI. That is the basis of what people were hired on before AI, because no alternative existed.
This is literally the recipe for the coming software apocalypse.
Just watch the software engineers who say "Let's be careful" lose their jobs and get replaced by non-technical vibe-coders churning out 10K lines of dirty insecure, unmaintainable code per day... Ticking time bomb.
I swear, good engineers are going to move to North Korea to monetize because of the amount and size of 'opportunities' this will create.
It seems so far-fetched but that's the direction it seems to be heading.
exactly how I feel. I mean let's face it, we were overpaid and underworked because we were considered magicians, but the magic is gone.
What I fear is that it's impossible for the untrained to see the difference between a highly-technical developer and the vibe-coder.
Software is so buggy nowadays (just look at Facebook, or Facebook Ads. Crazy) I wonder what's going to happen in the long run.
I was banking on the opportunities it will create but nah, have you tried debugging AI-generated codes of thousands of lines where the original dev doesn't exist? It's a mess I don't wanna clean up.
>> I was banking on the opportunities it will create but nah
Seems we're on the same wavelength because I had the exact same thought about this.
But actually, I'm probably not so fussy. I wouldn't mind cleaning up the mess if I'm paid hourly... But now my concern is: What if they want me to clean up the mess but demand that I do it in a particular way which makes it impossible?
I'm already seeing signs of this. This was already kind of the case in my last job; I had to fix things and implement new features but we had to keep the same clunky, over-engineered architecture. My current job doesn't have the same degree of architectural legacy baggage but there is a lot of bureaucracy to deal with instead, which is itself restrictive.
Man, I would straight up tell them I will use AI to review that crap and so be it. In fact this is what I tell to colleagues if they send my non-tested vibe-coded stuff or vibe-generated documentation. Everyone can generate shit today, but I believe testing, reviewing, writing docs, etc. must still go through humans.
I'm not sure we yet have the measure of how much agents/chatbots will alter our cultural norms, communication style, etc..
A lot of very quick "AI is not the problem" responses (and all in their own top threads. writing rather than conversing.. perhaps a pattern there? but I digress)
Agree with the pretext that some problems may be due to bad habits being amplified
But to me that default chorus of the "you're holding it wrong" thought-terminating cliche sounds no different to
"social media is not an issue, people just need to use control"
or "ads are not the issue here, just ignore them.."
Or "hey, the button which randomly gives the pigeon seeds when pushed is not the issue here, pigeons just shouldn't be so needy"
Also I've seen good engineers (and non-engineers) get swept up in the "magic" of genai progressively lose chunks of their formerly consistent, and long-standing good behaviours
In my opinion, software engineering was never mainly about programming.
Programming is a rather small part of the software engineering cycle.
Programming by hand is one way (it also works with AI), to make decisions. Decisions about scalability, maintainability and sanity of your software.
And still this will pay off!
To me it sounds as if he is a very bad software engineer.
Software engineers are splitting into two camps either by choice or by being forced to by management. Those who have the self-discipline to use LLMs responsibly and those that are going all-in on AI. Imo there's a place for both camps but I don't think they can coexist in the same team.
> Well, as much as people lose all their muscle by been bed-ridden, he's lost all his brain by being AI Assisted™.
> THIS is what I am so against when people use AI for everything.
This sounds horrendous but can we please stop equating vibe-coding to AI-assisted coding? They are not the same thing. It's one thing to submit 5K vibe-coded untested PR and another thing to use AI to add functionality in a granular way that you can review, or generate boilerplate, or test your ideas and scrap them if they suck and so on.
Thanks for expanding on this! I appreciate it and my sympathy to you.
I do have a similar case at work, but it's my superior in that case - shipping large volumes of code with Claude, 100% vibed. It's slightly different in the sense of him having almost no clue about programming but having very well self-organization and discipline. Rigorously documents his work (with Claude of course) and commits his Claude sessions in the repo. Funny enough we got 2 paying customers for one the projects he vibed. It blew my mind. I don't know how this will play out, but it's been a bit scary and pushed a lot of us to do more with AI rather than manual coding. I still can't get myself to commit to 100% vibe-coded stuff though.
> After this project, I'm done. I might turn into a yoga teacher.
I gather that you don't have the "juice" to demand better. Even AI testing AI would be better (in some parallel universe?) than not testing at all.
So become a consultant, preaching the simple gospel of "test this goddamn thing before unleashing it, or the tougher gospel of "Well if AI is doing all the work, then why don't you find a job that tolerates garbage?"
I almost always had more responsibility than authority, except for a job where I actually had a field team. They were laid off by the "authorities" in order of greatest competence first and I ended up with one guy who made countless goofs, like creating overlapping disk partitions. I ended up redoing all his work, more than doubling my load. Does that sound familiar?
In your case, it sounds like you will end up similarly, unless you find a way get your management to establish a "No broken shit" rule. I know that's daunting, and in many jobs, I could only finagle minor changes, usually by talking with other people more tenured, who had a piece of management's ear.
If it amounts to "I must double my work or this all falls apart" while your partner pushes buttons (my charge was a "button-pusher"), then leave, and good luck.
But do give it one college try to get some sanity into the system.
I did work as an Engineering Manager for a bit but honestly hated it with my guts. Promised myself I would never do that again. I have two young children so they're a bigger priority right now.
Not to sound pretentious but I was top 10 in my country (for example, I read a 500-page book on Assembly language the night before and aced a test that 90% of people failed) so it's really hard for me to work with people who don't get things as fast as I do.
Nothing to brag as it really isn't all hard work, I'm just quite neurodivergent compared to the average. It comes with A LOT of downsides (some nights I only sleep 2 hours because I came up with the solution to a hard problem while I was asleep and wake up to code it and it works)
Now, all this is because I am now at a crossroad where I have enough money and life experience in other jobs (I stopped programming for 10 years in fact because I had a pretty successful business).
The sad part is that I rekindled my love with programming in the last decade (I'm past 40), but it might be time to let it go and go back to doing business.
EDIT: I did work as a consultant (reluctantly) and was charging $500/hr because I really didn't like it but the demand was there. With the economy not being what it was and all the "AI prompt vibe coders" etc blah blah out there I don't really know what's a good rate now.
I don't know about "coding from scratch". Everything seems to have gone up one notch, and I don't really know what that implies.
I'm a real old-timer and really wanted that kind of leap for a long time, and delved into scripting languages and even dataflow as in the old Prograph [0], but did mostly sysadmin kinds of chores, with some very fun scientific simulations (so that upper management, even older than I was) could use them, in Hypercard, and that dead-ended.
If the next step is to build upon "building blocks" I'd like to see THOSE blocks thoroughly tested, as math libraries are to a great degree, and not ad-hoc. The most annoying thing I find about LLM's is their propensity to reinvent wheels every time. This may be inherent in the design, I don't know. But a serious business or other project should insist on reliable reusable blocks.
Perhaps the current situation is more like rounding up animals after they broke out of the zoo than caring for them, before it stabilizes if it ever does.
I nag people here to meet some real unmet needs, particularly for seniors, not just because I am one, but because I have had a lot of ideas on what to do, especially about building community and services. The internet is just too savage a place for most. I'll bet that the people approaching senior status will bitch about the lack of real self-driving cars, exoskeletons, and fraud-proof networks of trust, compared to the chainsaw internet of scams --- that THEY could have solved.
Create a "programmable internet" to get around all the annoyances, traps and dark patterns? I do it all the time, working around annoyances, popups, scams, paywalls. The average person isn't going to be able to do these things, not even with the leveling up of programming, but perhaps via proxy. Just as everyone seems to be putting up a shield, with endless CAPTCHAS (and I have browser extentions to solve these relentless eye exams) maybe users need a shield themselves.
But I ramble.
Good luck in whatever you decide to do.
Tally your strengths and strongest interests. Enjoy your family. I was posting elsewhere about some of the insane managers I've endured but memories and photos (some scanned from the film days) of being a parent are the real heart and soul of life. I have read some really insightful articles on what skills will be needed in the next few years and I'll summarize as I make sense of them.
yeah that's exactly my worry. I'm sure there could be some opportunities if I were in the US but I'm neither a citizen nor a resident. I've worked remotely my whole life and not looking to relocate.
> Well, as much as people lose all their muscle by been bed-ridden, he's lost all his brain by being AI Assisted™.
Wanna know the answer of those who lost all their brain cells to that?
"If there are bugs, it's only because the prompt was off. I'm a better prompt-engineer than your business partner, so this could never ever happen to me."
And off to the couch they go sit and shut their brains off.
used to respect my skills, since we started this project before AI, but as I told him in the beginning, the more he asks a sycophantic AI stuff, while asking me less, the less he will respect my opinion. It's just the nature of things.
You like those who agree with you more than those who disagree, but real-life learning doesn't come from everyone telling you you are correct all the time.
I believe we're at the point of no return where he didn't ask me a single question before sending a 5K LOC PR
"no worries man I know what I'm doing" it’s first sentence in AI-using human mouth that tells me to turn around and run.
I’m using models, I’ve built two whole projects without looking at more than 1% of code, but I’ve tested the interface and went through data on disk to verify the assumptions.
Sad to see you have a FRIEND who behaved this way.
> I’m using models, I’ve built two whole projects without looking at more than 1% of code, but I’ve tested the interface and went through data on disk to verify the assumptions.
so, would you say "no worries man I know what I'm doing" about how you use AI?
The problem is not vibe coding. The problem is the people doing this were never good software engineers before vibe coding.
Simply put you still need good practices to produce working good code with AI.
Oddly what I have found best is patterns that many would cringe and demeans unit test and integration test.
Yes I still vibe unit test and integration test, but more than ever I have to put on my PM and QA hat, even more so with web dev.
For me the only part of the process that has changed is I don’t write the code. I still setup environments to run the code, I still manually poke at things I suspect will be fragile. But most importantly I still always run the software before shipping it.
> Yes I still vibe unit test and integration test,
False sense of security. What's stopping the agents from doing an elaborate form of "Assert.IsTrue(true)" in some/most/all of the tests since you are vibing it and not reviewing it?
Exactly what happened in a PR from a few months ago, from the beginning it felt wrong, pytest was moved from the dev group to the main import section. Then imported in our models. To add ignores everywhere.
Except that it is the enabler. Without AI there are no 5K lines of code that no human has ever seen. AI introduces a new category of problem that did not exist before.
Couldn't have said it better myself. Issue at hand is if you're working in an environment that enables such people and even encourages that behaviour. Suddenly, their voice is the same where it was never even a whisper before, and you can't do anything about it. It is extremely demotivating and I would actually propose a more radical approach to it - either embrace the shit in full and commit to it as well or really do engage into another company where there's a better filter for hiring. The latter are usually smaller companies because they can't afford no to.
Well said. The problem is that vibe coding can paper over bad engineering practices for awhile. And results look good, and when you raise concerns people look at you like you're dumb and trying to slow things down.
But the underlying bad practices _will_ bite the same they would have before AI, you just don't know when.
It's a really frustrating state of the industry to be in.
Doesn't sound like AI is necessarily the root the problem. If he's not the person to test, iterate, and deliver quality with AI, he probably wouldn't be without it. You just figured it out sooner.
Or, if you have reason to believe he's got what it takes - treat it like an leaving opportunity, give him room to make mistakes. This is, ime, what a rookie AI user mistake looks like.
It is easy to get carried away by the momentum. Many have fallen into that trap, not all are irredeemably bad software engineers.
> If he's not the person to test, iterate, and deliver quality with AI, he probably wouldn't be without it
I see the take quite often on HN, and I have a major gripe with it. Yes, people were lazy and we had these problems before LLMs hit the scene, but clearly there has been a sea-change in both the frequency and degree of this behaviour since.
A problem can both exist independent of a catalyst and be exacerbated by it. Arguments that try to window-dress this away are not particularly productive.
It's a sea change because a lot of this code looks plausible and seems to work to someone inexperienced at assuring quality, and it will take some time before the people emboldened by that learn.
I just yesterday had an agent produce a prototype that appeared to work and that looked plausibly well engineered, except I knew from experience that what it had done should be doable with a tenth of the codebase.
Someone without experience might've just accepted that prototype. It took one prompt to get it turned into a reasonable MVP, but you'd need to know what to ask, and you'd need to know you needed to ask.
I agree with your sentiment. Simply saying 'AI' is not the reason, clouds what is actually happening. To illustrate I suggest playing the word swapping game. Swap AI in the sentence "AI isn't the problem" with, guns, drugs, gambling, or dare I say social media!
Human behaviour is complex. Patterns of behaviour that occur in general (there are always outliers of course) dictate so much in society, culture, politics and inter-human relations. Changing these behavioural patterns will affect meaningful real world change.
Social media has affected change in human behaviour. I suspect AI will too. We are starting to see the first shoots of the changes AI will bring.
Human behaviour on the average will choose convenience over effort. It is very difficult to buck this trend societally. To be clear, AI provides an immense amount of convenience.
I think the argument is quite valid. OP writes "business partner". It begs the question - would their business partner get on board if they didn't have a AI subscription by their side? How about instead of 5K PR, OP got the same half-baked buggy crap or absolutely nothing instead? We don't know the business partner, but we are ready to throw the problem at AI agents right away. Well, I find this not particularly productive.
> but we are ready to throw the problem at AI agents right away
I'm not sure I understand this correctly, but I assume it means blaming the agents? If so, that's not my point - I'm describing how many people are using them to be lazy, which shouldn't be particularly controversial.
To your other question: if OP had gotten half the size PR to go through, that sounds more manageable to me? If they had gotten nothing, that could be a useful signal about the capability of their business partner, worth knowing now rather than later. Could one imagine a situation where AI enables a thoughtful business partner to provide more value? Absolutely. Should this be the default assumption? I'd argue that it absolutely shouldn't.
we started this project before was mainstream (we got sidetracked by other work) and him having less experience but being a very good programmer (I went with him to college and worked together for years), he would listen to my input, fix, and adjust. I have decades of experience.
This is all but gone
Really sorry to hear so. Your story reinforces a fear I have that humanity is being squeezed out of all manner of things (pre AI too, but even more so since).
Well, I’d say LLMs scale up the underlying starting point.
It’s still shocking to me pre-AI how often people would ship code they didn’t bother to run once, relying on some QA stage to report an issue.
With AI the solution to it is also trivial since an extra three seconds spent writing a prompt could have had the agent run and verify it.
> With AI the solution to it is also trivial since an extra three seconds spent writing a prompt could have had the agent run and verify it.
This just recurses the original problem by an additional step: assuming someone is even bothered to write the prompt, they are just as unlikely as before to check what it spits out.
AI may be a productivity amplifier for some, but I don't think it's far-fetched to describe it as a laziness magnifier for the majority.
*learning opportunity. Damn autotype
There is a certain irony blaming the autotype rather than the lack of review...
AI is not the problem here, but the difference in deep beliefs of the company founders.
When I had partners in my business I studied them first deeply like I was going to marry them because I WAS going to marry them!
You should agree on your values or else the company will go nowhere. One founder will raw one way and the other the opposite making the boat to stay in the same place.
I have a software company and the opposite problem: I am obsessive about testing and meta programming. We write way more testing code than code, we automate everything with meta programming.
It was clear several times that we tested and automated way more than necessary. We automated things that only used once for example(although we used later multiple times on other projects).
So we had not the usual problems than companies have like bloat or security vulnerabilities but we had the problem of having software that was was super expensive because we did double or triple the code than the competition with tests that developers hated(they are boring to write). Our software took also way more time to market.
Companies love to buy the cheapest they can today and tomorrow we will see, so that put us in a difficult position. It was specially hard at first, but then this principles gave us an edge(because most companies would compromise early and not reap the benefits later).
Had I a business partner with a different mindset and beliefs, it would have been impossible to me not to change course of the boat in hard times.
BTW we use LLMs with great success. It let's us test even more spending way less than before.
Thanks for your input. I find testing (especially FE, where a single change can ripple through the whole chain) is actually a great way to use LLMs
It's sad because I've known this guy for 20 years and since the beginning he had a tendency towards "work smarter, not harder", a good motto at the time you're 20 but I often saw it crippling him in the long run, while having short-term gains.
Long-term, funny enough, he said he wanted back into this project because he was very unsatisfied with "easy money" (we both got into SEO at the right time and made a ton of easy money) and wants something stimulating for his brain.
In the meantime I had kids, worked my ass off to get the health and fitness level I want (he also wants it... alas nothing changes).
I'm kind of quite committed to this and we're maybe 30% away from finishing, but I almost debate sometimes if I should just do the FE vibe-coded and let him see that it won't be as good as I can manually do it (FE is hard in the end). Just get it out of the door, keep the customers that we have and let it kind of die down.
OP and long time lurker here, I'm desperate and looking for honest-to-god advice on what to do.
I am one of those neurodivergent people who loves programming because it gives me the right stimulation.
I also enjoy drawing, gardening, and exercise.
I don't own a car because I love walking, and such.
So I'm wondering if it's time to call it quits with programming (for a living).
Luckily I have some savings and I have other interests from years of doing random jobs and businesses, and as much as I love programming, especially with others, I think very few of us still want to program.
I'm like the Last Samurai of programming while everyone is using AI-assisted gattling guns.
Is it time to let go?
The differentiator right now is taste. Most people have none (like the business partner in the OP), and some people have plenty (like you seem to have). These two sets of people will produce wildly different results with the same tools.
Two things you can do regarding programming here.
1. Be a consultant for a high wage, use AI, produce good results quickly - little to negative intellectual joy here. You still have the edge here due to good taste.
2. Produce 'artisanal' code in your spare time. Work 3 months for pay, then 9 months on whatever projects scratch your itch.
I still want to program and I don't think it's time to let go. Hit me up (email in bio) if you wanna chat, you sound like a really interesting bloke.
1. is a bit contradictory or at the least a "devils in the details" wishful fantasy which hasn't been showcased convincingly in the wild yet (by "the wild", I mean software eng shops and corps). It's assumed the devs who have taste were able to implement their taste fine pre-AI, sure, but the whole gotcha in 1. is that the ones with good taste currently are not confident in implementing their good taste while simultaneously "producing good results quickly using AI".
I also really enjoy programming, and relate to your comment below about sometimes waking up in the middle of the night because your brain figured out how to code a hard problem.
Sorry to hear you're having a rough time with work. I do think AI availability has made otherwise smart people put out lazy work. But I also think that there are more people who still "code by hand" than you think.
You might find some joy by going to certain events where people talk about different items coding-wise. I enjoyed the talks for this conference recently https://www.youtube.com/@softwareshouldwork (particularly Richard Hipp's, Delaney Gillilan's and Carson Gross' talks), maybe you can find similar spaces. The benefit has been Youtube's algorithm recommending me better software talks now too.
I think AI is not really the subject. You probably chose a bad partner. As for quitting, since you mentioned gardening:
"C’est une folie de haïr toutes les roses parce qu’une épine vous a piqué."
I’ve been on a sabbatical for 99% of this year, small pockets of time spent doing some maintenance for clients I love.
I bought a seriously derelict property in country Australia and am having a good time slowly fixing it up.
I’ll get back to programming next year, hopefully the blow-ins have moved on to the next hype cycle by that time.
> who loves programming because it gives me the right stimulation
A small (and a bit tangential) personal story: I thought I am one of those people who care about the result, more then about the process.
But, I've recently played a game called "The Farmer Was Replaced" that someone mentioned in the comments in one of the posts here, on HN.
The game is basically a farming sim, except you farm by programming a drone using a Python-like language.
Of course, I didn't use AI, because where is the fan in this? In "normal life" (work, personal projects), it is very hard for me to convince myself to do anything manually, because generating code is more efficient. But in a game, using AI assistance would be like using aimbot in a shooter or a cheatcode in some other game. Yeah, it's much more efficient, but what is the point? There is no challenge and no fun that comes from overcoming the challenge.
So, anyway, I was writing code by hand for the first time in I-don't-even-remember-how-many years. And it was so much fun. I don't remember when was the last time I felt so engaged, so immersed, so "in the flow".
So, yeah. After all, maybe I'm not a ruthless pragmatic who only cares about the final result and doesn't gives a shit, how did he get there...
I booted my MiSTer into TI-99/4A mode a week or so ago and started writing little programs in TI BASIC. TI BASIC sucked. It was slow, due to being written in a form of bytecode which had to be interpreted (and it would in turn interpret your program). And you couldn't do sprites, not without getting the Extended BASIC cart.
But HOLY SHIT THE DOPAMINE RUSH. Just crafting things and making them appear on the screen with a few statements... It was like I was a kid again. I remembered why I liked programming.
It's like... AI is a cheatcode. It's iddqd. Sure, you can cakewalk through a level with it on, but are you really playing the game, then? You certainly haven't gotten better by doing so, and you haven't experienced the involuntary buttcheek clench when you hear a cacodemon hiss from behind you... The whole point of programming is the new things I can do by having gone through the exercise of figuring it out. The ways I've grown, the glimpses I've gotten into the world.
You are far from alone on that path and indeed I also wonder if it may be time to let it go.
Just find a better place to go. If you have savings, quit this nonsense and start looking for something else while working on yourself.
I just got back to working on a project I like where we don't use any LLMs for code generation because we think it kind of sucks and leads to bad results unless it's for incredibly small contexts (which is usually not worth the prompting).
You're also not as alone as you think in actually liking programming enough to do it by hand even after having used LLMs. There's also the fact that you build more theory up by actually writing the code anyway, so someone who actually wants to know their code base intimitely will always have to write the code.
Why not become a machine gun operator?
No it's absolutely not time to let go. You are one of a limited set of people with pre-AI taste and intuition for good software. They aren't making any more of them. And you'll be a super valuable asset anywhere you go.
I mean, if you want to retire go ahead, but don't do it because of AI. AI is just a tool to generate code faster. It doesn't replace the work of actually making design decisions and knowing what will scale and what's better going forward vs what will work now but spiral into unmanageable spaghetti code soon (yes AIs also have a harder time managing messy code than neat code)
I often read here comments along the lines of "there will be plenty of jobs for AI-unfuckers" [1], but I fear the truth may be closer to "sometimes things are broken, are then fixed with a broken solution, and will remain broken till the end of time".
In my last round of job search this year no one asked me questions about good software practices but they did ask me how proficient I was with AI agents. And let's not forget that we are part of an industry whose motto is "move fast and break things". Sure, NASA has a zero-bugs approach to software dev, but they don't have room for all of us.
[1] https://news.ycombinator.com/item?id=44646820
I'm am certainly an "ai believer" and have been since I saw the tragectory we were on.
It was pretty simple to project the compounding effects and headroom that we had. I go back and forth on this all the time.
There probably won't be much demand for "un-fuckers". Up until recently I thought knowing what to build for a given context would give you quite a bit of runway.
I still think knowing what you should build will be important for the next 12-24 months.. but I have to be honest.. after that.. I don't know.
I don't know how to feel myself. I'm certainly an "ideas" guy as well as a nitty gritty build it guy. I just think my technical side is depreciating very rapidly.
Completely agree. Saying this as someone who loves handcrafting code and has been doing it professionally for decades, it looks like a dead end at this point. LLM generated code might not be pretty, but these damn models are already so good at handling complexity, that they can actually work fine with a byzantine code base (which honestly would also describe most human-derived codebases I have worked with). I am still reeling from this realization and trying to figure out what it means.
SV echo chamber here it seems. No, most businesses globally don't work in idiotic and often amoral mode of 'move fast and break things' (and unlawful on top of that, but who has time for some boring stuff like that when greed fully takes over).
Banks will not work like that for next 10 years at least. The tiny amount of money one could save on those big banking projects by using llms aggressively will bite back immediately when shit just doesn't work, clients leave for good and regulators will fine you to the ground (or remove your banking license). Many other businesses who are not race to the bottom and realize how crucial reliable IT is to their business also won't go that way anytime soon.
World is not ending, there is some transformation and thats about it. Its true there will be much less jobs for devs but its mostly affecting newcomers. Statistical models are not making code better than experienced humans, and won't for some time if ever.
> most businesses globally don't work in idiotic and often amoral mode of 'move fast and break things'
Not sure what business you are in but the non tech industries (corporate IT, in house apps) have not, in general, been associated with quality. Cost cutting, outsourcing to the lowest bidder: yes. Quality: no.
Mostly banking but dipped my feet in many other areas in the past (telco, insurance, municipalities, army etc).
Its true budgets are often the problem, but not much 'move fast and break things' mantra of SV, rather 'move as much as low budget and corresponding quality and bureaucracy allows you to'
Look at the COBOL devs. They are been called in their retirement years. You don't want that to be you. Life is short
Have you seen what they charge per hour? Let’s just say it’s not Upwork rates.
This is what’s know as “cope”. People are not being hired on the basis of how well they know how to not use AI. That is the basis of what people were hired on before AI, because no alternative existed.
This is literally the recipe for the coming software apocalypse.
Just watch the software engineers who say "Let's be careful" lose their jobs and get replaced by non-technical vibe-coders churning out 10K lines of dirty insecure, unmaintainable code per day... Ticking time bomb.
I swear, good engineers are going to move to North Korea to monetize because of the amount and size of 'opportunities' this will create.
It seems so far-fetched but that's the direction it seems to be heading.
exactly how I feel. I mean let's face it, we were overpaid and underworked because we were considered magicians, but the magic is gone.
What I fear is that it's impossible for the untrained to see the difference between a highly-technical developer and the vibe-coder.
Software is so buggy nowadays (just look at Facebook, or Facebook Ads. Crazy) I wonder what's going to happen in the long run.
I was banking on the opportunities it will create but nah, have you tried debugging AI-generated codes of thousands of lines where the original dev doesn't exist? It's a mess I don't wanna clean up.
>> I was banking on the opportunities it will create but nah
Seems we're on the same wavelength because I had the exact same thought about this.
But actually, I'm probably not so fussy. I wouldn't mind cleaning up the mess if I'm paid hourly... But now my concern is: What if they want me to clean up the mess but demand that I do it in a particular way which makes it impossible?
I'm already seeing signs of this. This was already kind of the case in my last job; I had to fix things and implement new features but we had to keep the same clunky, over-engineered architecture. My current job doesn't have the same degree of architectural legacy baggage but there is a lot of bureaucracy to deal with instead, which is itself restrictive.
The only proper response is a “LGTM” and merge it into main.
5k is rookie numbers. My PM sent me an 80k line PR and told me to yolo it into production last week.
Man, I would straight up tell them I will use AI to review that crap and so be it. In fact this is what I tell to colleagues if they send my non-tested vibe-coded stuff or vibe-generated documentation. Everyone can generate shit today, but I believe testing, reviewing, writing docs, etc. must still go through humans.
haha seriously? Damn I really gotta buckle up
I'm not sure we yet have the measure of how much agents/chatbots will alter our cultural norms, communication style, etc..
A lot of very quick "AI is not the problem" responses (and all in their own top threads. writing rather than conversing.. perhaps a pattern there? but I digress)
Agree with the pretext that some problems may be due to bad habits being amplified
But to me that default chorus of the "you're holding it wrong" thought-terminating cliche sounds no different to
"social media is not an issue, people just need to use control" or "ads are not the issue here, just ignore them.."
Or "hey, the button which randomly gives the pigeon seeds when pushed is not the issue here, pigeons just shouldn't be so needy"
Also I've seen good engineers (and non-engineers) get swept up in the "magic" of genai progressively lose chunks of their formerly consistent, and long-standing good behaviours
In my opinion, software engineering was never mainly about programming. Programming is a rather small part of the software engineering cycle.
Programming by hand is one way (it also works with AI), to make decisions. Decisions about scalability, maintainability and sanity of your software. And still this will pay off!
To me it sounds as if he is a very bad software engineer.
Software engineers are splitting into two camps either by choice or by being forced to by management. Those who have the self-discipline to use LLMs responsibly and those that are going all-in on AI. Imo there's a place for both camps but I don't think they can coexist in the same team.
> Well, as much as people lose all their muscle by been bed-ridden, he's lost all his brain by being AI Assisted™.
> THIS is what I am so against when people use AI for everything.
This sounds horrendous but can we please stop equating vibe-coding to AI-assisted coding? They are not the same thing. It's one thing to submit 5K vibe-coded untested PR and another thing to use AI to add functionality in a granular way that you can review, or generate boilerplate, or test your ideas and scrap them if they suck and so on.
That's the thing, that's how he started, but knowing his personality, I knew the temptation would be too high. I warned him.
Then yesterday when I started asking him questions about how certain stuff worked, he had absolutely no clue.
He didn't even read the docs summarizing what the PR did.
In his mind AI is just assisting him. It's just like the alcoholic saying "one more drink won't harm me, I can quit any time".
Thanks for expanding on this! I appreciate it and my sympathy to you.
I do have a similar case at work, but it's my superior in that case - shipping large volumes of code with Claude, 100% vibed. It's slightly different in the sense of him having almost no clue about programming but having very well self-organization and discipline. Rigorously documents his work (with Claude of course) and commits his Claude sessions in the repo. Funny enough we got 2 paying customers for one the projects he vibed. It blew my mind. I don't know how this will play out, but it's been a bit scary and pushed a lot of us to do more with AI rather than manual coding. I still can't get myself to commit to 100% vibe-coded stuff though.
> After this project, I'm done. I might turn into a yoga teacher.
I gather that you don't have the "juice" to demand better. Even AI testing AI would be better (in some parallel universe?) than not testing at all.
So become a consultant, preaching the simple gospel of "test this goddamn thing before unleashing it, or the tougher gospel of "Well if AI is doing all the work, then why don't you find a job that tolerates garbage?"
I almost always had more responsibility than authority, except for a job where I actually had a field team. They were laid off by the "authorities" in order of greatest competence first and I ended up with one guy who made countless goofs, like creating overlapping disk partitions. I ended up redoing all his work, more than doubling my load. Does that sound familiar?
In your case, it sounds like you will end up similarly, unless you find a way get your management to establish a "No broken shit" rule. I know that's daunting, and in many jobs, I could only finagle minor changes, usually by talking with other people more tenured, who had a piece of management's ear.
If it amounts to "I must double my work or this all falls apart" while your partner pushes buttons (my charge was a "button-pusher"), then leave, and good luck.
But do give it one college try to get some sanity into the system.
Thanks for your reply.
I did work as an Engineering Manager for a bit but honestly hated it with my guts. Promised myself I would never do that again. I have two young children so they're a bigger priority right now.
Not to sound pretentious but I was top 10 in my country (for example, I read a 500-page book on Assembly language the night before and aced a test that 90% of people failed) so it's really hard for me to work with people who don't get things as fast as I do.
Nothing to brag as it really isn't all hard work, I'm just quite neurodivergent compared to the average. It comes with A LOT of downsides (some nights I only sleep 2 hours because I came up with the solution to a hard problem while I was asleep and wake up to code it and it works)
Now, all this is because I am now at a crossroad where I have enough money and life experience in other jobs (I stopped programming for 10 years in fact because I had a pretty successful business).
The sad part is that I rekindled my love with programming in the last decade (I'm past 40), but it might be time to let it go and go back to doing business.
EDIT: I did work as a consultant (reluctantly) and was charging $500/hr because I really didn't like it but the demand was there. With the economy not being what it was and all the "AI prompt vibe coders" etc blah blah out there I don't really know what's a good rate now.
I don't know about "coding from scratch". Everything seems to have gone up one notch, and I don't really know what that implies.
I'm a real old-timer and really wanted that kind of leap for a long time, and delved into scripting languages and even dataflow as in the old Prograph [0], but did mostly sysadmin kinds of chores, with some very fun scientific simulations (so that upper management, even older than I was) could use them, in Hypercard, and that dead-ended.
If the next step is to build upon "building blocks" I'd like to see THOSE blocks thoroughly tested, as math libraries are to a great degree, and not ad-hoc. The most annoying thing I find about LLM's is their propensity to reinvent wheels every time. This may be inherent in the design, I don't know. But a serious business or other project should insist on reliable reusable blocks.
Perhaps the current situation is more like rounding up animals after they broke out of the zoo than caring for them, before it stabilizes if it ever does.
I nag people here to meet some real unmet needs, particularly for seniors, not just because I am one, but because I have had a lot of ideas on what to do, especially about building community and services. The internet is just too savage a place for most. I'll bet that the people approaching senior status will bitch about the lack of real self-driving cars, exoskeletons, and fraud-proof networks of trust, compared to the chainsaw internet of scams --- that THEY could have solved.
Create a "programmable internet" to get around all the annoyances, traps and dark patterns? I do it all the time, working around annoyances, popups, scams, paywalls. The average person isn't going to be able to do these things, not even with the leveling up of programming, but perhaps via proxy. Just as everyone seems to be putting up a shield, with endless CAPTCHAS (and I have browser extentions to solve these relentless eye exams) maybe users need a shield themselves.
But I ramble.
Good luck in whatever you decide to do.
Tally your strengths and strongest interests. Enjoy your family. I was posting elsewhere about some of the insane managers I've endured but memories and photos (some scanned from the film days) of being a parent are the real heart and soul of life. I have read some really insightful articles on what skills will be needed in the next few years and I'll summarize as I make sense of them.
[0] https://en.wikipedia.org/wiki/Prograph
Which business will hire a consultant who doesn't fully endorse the AI-everything religion?
Which businesses? Ones who have refunded $10,000 on a $5.00 order or whose database was corrupted by sloppy code.
Unfortunately, there is demand for people who know to sanity check things. It's not glamorous, but neither is damage control.
yeah that's exactly my worry. I'm sure there could be some opportunities if I were in the US but I'm neither a citizen nor a resident. I've worked remotely my whole life and not looking to relocate.
> Well, as much as people lose all their muscle by been bed-ridden, he's lost all his brain by being AI Assisted™.
Wanna know the answer of those who lost all their brain cells to that?
"If there are bugs, it's only because the prompt was off. I'm a better prompt-engineer than your business partner, so this could never ever happen to me."
And off to the couch they go sit and shut their brains off.
Classic Dunning-Kruger, do they show respect for your skills?
used to respect my skills, since we started this project before AI, but as I told him in the beginning, the more he asks a sycophantic AI stuff, while asking me less, the less he will respect my opinion. It's just the nature of things.
You like those who agree with you more than those who disagree, but real-life learning doesn't come from everyone telling you you are correct all the time.
I believe we're at the point of no return where he didn't ask me a single question before sending a 5K LOC PR
This is enforced in my company so buckle up
[dead]
"no worries man I know what I'm doing" it’s first sentence in AI-using human mouth that tells me to turn around and run. I’m using models, I’ve built two whole projects without looking at more than 1% of code, but I’ve tested the interface and went through data on disk to verify the assumptions. Sad to see you have a FRIEND who behaved this way.
> I’m using models, I’ve built two whole projects without looking at more than 1% of code, but I’ve tested the interface and went through data on disk to verify the assumptions.
so, would you say "no worries man I know what I'm doing" about how you use AI?
The problem is not vibe coding. The problem is the people doing this were never good software engineers before vibe coding.
Simply put you still need good practices to produce working good code with AI.
Oddly what I have found best is patterns that many would cringe and demeans unit test and integration test.
Yes I still vibe unit test and integration test, but more than ever I have to put on my PM and QA hat, even more so with web dev.
For me the only part of the process that has changed is I don’t write the code. I still setup environments to run the code, I still manually poke at things I suspect will be fragile. But most importantly I still always run the software before shipping it.
> Yes I still vibe unit test and integration test,
False sense of security. What's stopping the agents from doing an elaborate form of "Assert.IsTrue(true)" in some/most/all of the tests since you are vibing it and not reviewing it?
Exactly what happened in a PR from a few months ago, from the beginning it felt wrong, pytest was moved from the dev group to the main import section. Then imported in our models. To add ignores everywhere.
If you're working with agents that severely misaligned you'll probably have issues way beyond fake tests.
It seems to me that chances of misalignment are a function of accumulated usage time. Eventually, it will happen.
> The problem is not vibe coding.
Except that it is the enabler. Without AI there are no 5K lines of code that no human has ever seen. AI introduces a new category of problem that did not exist before.
Couldn't have said it better myself. Issue at hand is if you're working in an environment that enables such people and even encourages that behaviour. Suddenly, their voice is the same where it was never even a whisper before, and you can't do anything about it. It is extremely demotivating and I would actually propose a more radical approach to it - either embrace the shit in full and commit to it as well or really do engage into another company where there's a better filter for hiring. The latter are usually smaller companies because they can't afford no to.
Well said. The problem is that vibe coding can paper over bad engineering practices for awhile. And results look good, and when you raise concerns people look at you like you're dumb and trying to slow things down.
But the underlying bad practices _will_ bite the same they would have before AI, you just don't know when.
It's a really frustrating state of the industry to be in.