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.
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.
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'
There's a lot of distortion on this forum. Some people don't consider anything less than being overpaid by venture capital furnaces as worthwhile work.
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.
You're misunderstanding. There is space for people who aren't so stupid that their prompts are stupid and their output is stupid and they sling and spam their stupid shit all over your system and attempt to get it committed and screw everything up.
> I often read here comments along the lines of "there will be plenty of jobs for AI-unfuckers"
That's not what I've said though.
> "sometimes things are broken, are then fixed with a broken solution, and will remain broken till the end of time".
That's true of a lot of software. But the big bucks things right now have too much money riding on them to be broken and remain broken. Every big tech co right now is talking about and sharing solutions for how to handle the wave of AI generated code that hits their systems from internal devs every day. Reliability is not at levels it was before. Change safety is far more challenging. Move fast and break things was shelved long ago at these companies. It's only popular in the public imagination.
That's not to say these companies will stop using AI. Absolutely not. It's a great tool, but like all tools it needs an ecosystem built around it. It needs evals, debugging, maintenance, telemetry, etc. It ultimately needs goals and directions driven by human beings who understand the domain well enough to steer. And that's why devs like OP are important now and will remain so for quite a while bc the rise of AI is the rise of many new and challenging engineering problems.
> 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
So what? How does this change anything I said? Being proficient with AI agents is far easier if you have the domain knowledge about what they should produce and far harder if you have no idea.
And anyways, what you're asked in interviews is not a mirror of the job. I see many people every day at work who are lauded and praised for using their knowledge to course correct and improve AI agents and other AI workflows. The idea that there will come a point where they work perfectly and don't need to be messed with, well that's also true of software and services. But it doesn't mean big tech companies with quite stable main services just stopped hiring people years and years ago
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.
> 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.
EXACTLY! The dopamine we get from doing stuff with mostly our own hands is unbelievable. But social media, junk food, and adult content should have taught us that cheap dopamine is addictive and dangerous.
Most of us will learn the hard way (or never learn)
The economics of it are such that writing code is, for humans, a handicraft to be indulged in out of idle curiosity. When it comes to actually useful software, for personal or industrial use, that is better off produced by machine; it's simply a waste of human time and energy to write it by hand.
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.
Thanks for the recommendation, I in fact couldn't sleep past 5am today and woke up to check an SQLite presentation on that channel. Pure gold!
I think that's how it is, we just gotta fine-tune the algorithms on our favor. I'm doing it slowly with Youtube (I'm into rock climbing recently, and it's been a fun ride on YT). My Twitter feed can actually be pretty amazing, even my IG one. I don't check them much often any more but they're still pretty good. No BS.
The trick is not to fall for the stupid videos with millions of views. I just recently learned about Mukbang or something like that (a weird trend where people binge eat on YT). Crazy
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.
Damn, would love to work somewhere where no LLMs are needed. I have a friend working for Nintendo stuff (I think he does something for their secondary software, IE sales) and he is not allowed to use LLMs. Happiest man on earth.
If I were in the States, where I have connections, people would hire me easily. But for my wife and kids, it's not really possible to emmigrate.
Remote is crazy. I heard from a friend who got 100s of applications on a remote job on a single day.
What I'm sad about is when others call us out for being a ludite or so.
But now I understand that these people don't understand that some of us just love programming.
I mean if I say I love gardening, nobody is going to tell me "Oh you're a stupid ludite, just buy a tractor and glyphosate"
Or when I mention my love for long walks in nature, imagine somebody telling you "that's so inefficient why don't you buy a segway, you'll be able to explore 10x the number of places without wasting any effort and time"
I've been on a 3 year sabbatical, and I spend most summers in the countryside with my kids, grilling meat, planting berries, going in nature for hours-long walks.
It's really a great life ain't it?
So you're planning to go back for work? I'm unfortunately on my last 12-months or so of runway so time to figure out stuff.
Did you have a lot of savings or something for this? I suppose the closest the rest of us can do is see if we can work part time and make use of longer weekends to go on exploratory trips, or paid holiday (if you work for a company) to go on exploratory trips, see the world, that sort of thing.
I must seek these out then! I avoid YouTube a lot as its noise I don't need in my life, and I hate the "facial reaction" / pointing thumbnails, or thumbnails with arrows, all designed to make you look.
I'd stay loose. We're in the middle of a big change but no-one knows what the end result will be. I suspect there will be a need for all kinds of programmers.
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 disagree. There is a lot of boilerplate to write and mundane things to check, which AI can speed up a lot. I'm not talking about 'software engineering' here, but about 'programming', which consists of a lot of write DTO->write rest interface->write interface->write simple select query->map response. The good taste here comes from the architectural design and the UI design.
The amount of 'software engineers' who couldn't design a table to be readable or a flow to be intuitive is off the charts. AI neither helps nor harms here, it just makes implementing good and bad decisions faster.
Since the discussion was related to the job security of a programmer with good taste, I'm trying to read your comment from that perspective.
If I am understanding you correctly, you are claiming that an AI will not design a table to be readable unless a programmer with good taste is prompting it. Did I understand you correctly?
If so, then yes we are going to arrive at different conclusions, because my claim is that the table made by the programmer with no taste will be acceptable to the client, hence good taste won't save the other programmer because good taste will require more guiding and interference by the programmer. And then the client will just go with the no taste implementation.
If the programmer with good taste truly does not lose any time implementing their solution when compared to the other programmer, as you claim, then yes the client will go with good taste since all else are equal.
I just find those assumptions hard to believe at this stage.
1. I did consulting for a bit, $500/hr haha I was so reluctant to do it. I do admit that SV was just pouring way too much money back then. The good old days. I don't live in the US, and remote opportunities have been vanishing quite quickly. In the US I would have no doubt I'd get crazy good offers, I have lots of people who respect me there. My friend and I even got offered high positions at Youtube way back when.
2. I quit 3 years ago and have done a lot of artisanal code. Lots of fun. But next year I'll need to figure out stuff, hence why I accepted this small contract (ends next month) and thinking if it's time to move on.
BTW I have a family to feed and my son eats like a husky even though he's only 45 pounds, so there's that. My daughter is a big fan of Argentinean shrimp and caviar. You get the point.
I called it quits about 8 years ago (I'm 55), mostly due to the fact that I couldn't get excited about anyone's "next big thing". Got into teaching, which was a new, rewarding skill to work on mastering. (I've still got a long way to go--what a challenging field!)
But of course I am teaching CS, and now it's just a nightmare of trying to figure out how to teach it. We have a lot of people saying we still need to teach the basics in addition to AI-assisted, and we have a lot of people saying to just teach vibe coding, and a lot of people saying just teach SE and no coding. And we're just trying to do the best we can for the students, while wondering if the field even has a sensible future.
(Anecdata: universally, my hired grads have told me they are incredibly grateful to have gotten through college before LLMs hit the scene.)
And it's a pisser, because teaching CS was great. I get to try to optimize learning and talk about what I love? Heaven.
I think I'm in this category of software engineer in terms of how much I love the work:
Requirements gathering: 5%
Functional spec: 5%
Design doc: 20%
Programming: 65%
Deployment: 5%
And engineers who were here:
Requirements gathering: 20%
Functional spec: 25%
Design doc: 25%
Programming: 5% (or 0%!)
Deployment: 25%
are having a much easier time with this transition. Unfortunately for me, programming is also the most fun to teach.
I have a lot of other interests outside of CS, one of which would actually pay money (GIS). Not sure how to get paid for dualsporting, hiking, bicycling, geology, visiting museums, writing (I do get paid grocery money for this), hand-coding, or reading. But I'm open to suggestion. :)
Hey beej71, I'm 41 and I love advice from people with more life experience than me.
I also love teaching (not particularly great at it, but not the worst) and taught SALSA dancing for a few years. The money was okayish (enough for myself but not to feed my children) but I LOVED being around ppl.
I wasn't talented in dancing but my talent in programming really made me happy (and well paid), so that kind of sucks. Definitely same as you, I'd even say programming can be my 90% of joy and effort (in a bad way sometimes, but I manage to talk myself into dialing it down)
I can't imagine the hell teachers are going through, especially coding, with AI.
Luckily I have a lot of weird interests and some can pay well (for example, I've had a decent success teaching people how to lose weight doing intermittent fasting, imagine the randomness of my interests)
What is your next plan?
I wasn't joking on becoming a yoga teacher, as cliché as it sounds, because I did it and quite successfully so (I was young, handsome, and single back then, and my classes were PACKED. I'm still handsome but the other two are gone. People paid cash so I had a lot of bills in my home lol)
> 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.
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.
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.
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.
I actually quit bigger companies and SV contracts because the amount of hype was way overblown. Nothing could not be AI, really. Think about how crazy it is that Jensen Huang is saying "AGI is here", seriously. I first thought malice, now I think it's just plain religious Zeal.
We're not a company, it's just me and my pal. I keep 50% of the product and profits and he pays me to do that (because I have way more expertise than him)
curious question where you work, the AI craze is not happening?
curious question where you work, the AI craze is not happening?
of course it is. There's no escape IMO unless, as you've said, go solo or with a small team. I just don't know of any company of a notable size that's not into it full force. It's both disgusting and amazing to look at really. Give it a few years and if it fizzles out I'm curious what will happen. If it turns great, I'm even more curious what will happen. Weird times.
Just "yesterday" we were all careful what and how we do stuff, you have now people who barely understand things slopping together OS and people use it blindly (Omarchy), and then shocked pikachu faces when giant gaping holes are noticed in it.. and people still do not care. Amazing trainwreck of a timeline, hah.
I'm seriously considering pivoting myself to security.
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.
It's like the casino, crypto gambling, or eating junk food. We're wired for that short-term gains and few of us can escape it. I understand that, just sad that the casino slot machine had to hit our craft.
Models are the worst now that they ever will be. Frontier models do a phenomenal job of employing good engineering practices and code style compared to where they were mere months ago. Unassisted programming is dead. Programming as a human activity will be on a timescale of years (along with a vast swath of other knowledge work).
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.
> 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 I got $1 for every time me or anyone I know prompted AI, got the wrong answer, and people claimed "you need good practices / get better at prompting" I'd be a millionaire
The problem is these companies literally say "NO NEED TO LEARN ANYTHING JUST VIBE". Ton of ads on social media and YT claiming that.
Like my mom doesn't even know what a PC, a Macbook or a Linux computer is and we're telling her she can make apps?
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.
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.
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).
> 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.
> 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.
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.
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 think one of my most repeated tropes on HN is bringing up that scale and degree matter. A problem of a kind is of a completely different kind when the scale is another order of magnitude.
A simple example: with picture editing software people have been able to create believable fake content for a long time, it did require skill which limited the scale, with genAI creating in seconds, by anyone, a much more plausible fake image than most people were able to before creates a whole other kind of problem that didn't exist before.
It's a tired argument to reduce issues to analogous which existed before and discard how the scale and degree of it changes the class of problem...
Seems like exactly the kind of distinction that would be a common refrain on HN, yet somehow on this topic it's gone out the window.
> A simple example: with picture editing software
Actually reminds me of a story a professor in a comsci ethics class once told about how the police in South Africa (during the apartheid years) would take photos of everyone protesting the regime at the university campus. Back then (this was the 80s) nobody really cared - what were they going to do, go through all the photos by hand?
He said "imagine how much more concerned people would be today". That was in 2011 or so.
I literally was earning 5x times more working for contracts and companies but got super sick of it, I was happy to work with one single guy (who PROMISED to keep AI at bay, it was our original part of the deal) for a huge paycut
Well that didn't last long. The temptation was too big for him too
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?
what does "verify the data on disk" even mean? I guess without knowing what the project is it's weird to think testing it + checking some files is enough, at least for me
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.
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.
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.
> 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.
My guy in fact started by blaming himself for prompting wrong, and yesterday he was "upset" that his friend Claudio missed some important documentation details (three API keys were needed and one I had no idea where to get, not mentioned once at all)
no worries, a lot of these assumptions exist mostly in my head as a frustrated dev.
and yeah luckily I ended up quitting just right when coding was getting easy with LLMs and even back then (3 years ago) I remember our PMs and bosses saying "look why does it take you a week? I did in in two hours!"
Got riddance to get out of those jobs, BUT I do agree that paying customers != shipping software. Getting paid customers is a great skill and it's the skill my friend has (he's an A class marketer, and not in a fishy or scammy way). The guy just understands marketting. That's why I'm working with him.
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
social media in an excellent example, alcohol, or even marihuana where it's pretty much legal to consume in my home country and a total pain in the ass for those of us who hate the smell.
Sucks that my children have to smell it, really different from my chilhood.
I don't know almost anyone who preferred having a childhood today (with so much social media and doomscrolling) as compared to the 80's and 90's, when it was just kids playing hide and seek outdoor, a few Michael Jackson home-recorded cassettes, and of course Sonic 2.
I am genuinely worried about skill-rot. My peers code and design skills are degrading so quickly. AI is making each feature harder to debug than the last. That's a skill problem, and if our skills weren't rotting we'd be able to fix it adeptly. But the process seems to ruin our cognitive tools.
This is the thing that people just simply don't see.
And for those who say "well why code and use your brain?", may I remind them that cities where people do a lot of walking are so much enjoyable, better for people mentally and physically than places where people use the "efficient and smart way" of transportation, that being a car.
I remember my cousins in the States picking me up by car and walking their dog by car, inside a gated community that was very enjoyable to walk. Pathetic to see.
(BTW both my cousins are severly overweight, 200+ pounds)
It seems to me the author is gradually accepting the fact that business people (and their grandparents) can now produce thousands lines of code.
This was not feasible a year ago.
Now, regarding the quality of the code, the edge cases etc, surely someone familiar with the failure modes can review and prompt the AI better. But that doesn't negate the fact that these people are producing huge amount of code using natural language without any programming background.
A skill that used to take many months (if not years to master) and require handcrafting the software line by line is now done in English with almost no understanding of programming primitives.
And I'd argue that if the models, harnesses and IDEs keep getting better then a layman learning few software best practices would outperform a team of senior developers writing code manually.
Therefore, the developers refusing to adapt and adjust will find themselves gradually in a position where a layman can easily outperform their economic output by a huge margins. And in my opinion, the author just found himself in this position, and still in denial of what is happening.
122 comments
[ 0.17 ms ] story [ 17.2 ms ] threadI 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?
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)
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
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.
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.
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'
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.
Not really. They're only as good as whatever they are prompted for. If you prompt them for the wrong thing, they'll do it. Great tools, human error
I'll bet we'll all still work in software in 24 months
Yeah I could do management and all that crap but I was Eng Manager for 5 months and hated it.
Some of us are like huskies, we need to run on the Siberian Permafrost in order to thrive, else we get depressed.
That's me.
That's not what I've said though.
> "sometimes things are broken, are then fixed with a broken solution, and will remain broken till the end of time".
That's true of a lot of software. But the big bucks things right now have too much money riding on them to be broken and remain broken. Every big tech co right now is talking about and sharing solutions for how to handle the wave of AI generated code that hits their systems from internal devs every day. Reliability is not at levels it was before. Change safety is far more challenging. Move fast and break things was shelved long ago at these companies. It's only popular in the public imagination.
That's not to say these companies will stop using AI. Absolutely not. It's a great tool, but like all tools it needs an ecosystem built around it. It needs evals, debugging, maintenance, telemetry, etc. It ultimately needs goals and directions driven by human beings who understand the domain well enough to steer. And that's why devs like OP are important now and will remain so for quite a while bc the rise of AI is the rise of many new and challenging engineering problems.
> 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
So what? How does this change anything I said? Being proficient with AI agents is far easier if you have the domain knowledge about what they should produce and far harder if you have no idea.
And anyways, what you're asked in interviews is not a mirror of the job. I see many people every day at work who are lauded and praised for using their knowledge to course correct and improve AI agents and other AI workflows. The idea that there will come a point where they work perfectly and don't need to be messed with, well that's also true of software and services. But it doesn't mean big tech companies with quite stable main services just stopped hiring people years and years ago
Oh look a thing I never said being refuted
You pretty clearly did say this.
I saw him starting at a monitor at midnight, back all hunched up and thought "nope, not what I envision"
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...
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.
Most of us will learn the hard way (or never learn)
Most of my friends would use cheat codes to beat games.
Me and my brother (who's a crazy amazing artist btw) would instead figure out how to get good enough in a game to not lose a single life.
Contra, Ninja Gaiden, Mario.
I could see that the amount of fulfillment we got was 100 times more than my friends who played the game like that and didn't struggle.
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 that's how it is, we just gotta fine-tune the algorithms on our favor. I'm doing it slowly with Youtube (I'm into rock climbing recently, and it's been a fun ride on YT). My Twitter feed can actually be pretty amazing, even my IG one. I don't check them much often any more but they're still pretty good. No BS.
The trick is not to fall for the stupid videos with millions of views. I just recently learned about Mukbang or something like that (a weird trend where people binge eat on YT). Crazy
I almost sold my soul to the devil doing AI assisted coding review for top dollar. That night I almost cried in my sleep
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.
If I were in the States, where I have connections, people would hire me easily. But for my wife and kids, it's not really possible to emmigrate.
Remote is crazy. I heard from a friend who got 100s of applications on a remote job on a single day.
What I'm sad about is when others call us out for being a ludite or so.
But now I understand that these people don't understand that some of us just love programming.
I mean if I say I love gardening, nobody is going to tell me "Oh you're a stupid ludite, just buy a tractor and glyphosate"
Or when I mention my love for long walks in nature, imagine somebody telling you "that's so inefficient why don't you buy a segway, you'll be able to explore 10x the number of places without wasting any effort and time"
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.
I've been on a 3 year sabbatical, and I spend most summers in the countryside with my kids, grilling meat, planting berries, going in nature for hours-long walks.
It's really a great life ain't it?
So you're planning to go back for work? I'm unfortunately on my last 12-months or so of runway so time to figure out stuff.
Yes, there’s no complaints from me :) Spent 4 months camping and now I’m still camping but fixing a property.
I’ll do something for money next year, who knows, I might find a good programming contract or I might be serving coffee at the local cafe!
My home town had super undervalued real estate for roughly a decade prior to 2023/4.
I bought another stinker of a property nobody wanted.
I’d save and look for under valued property, my understanding is the US has ridiculous amounts of it provided you’re happy to live in small places.
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.
The amount of 'software engineers' who couldn't design a table to be readable or a flow to be intuitive is off the charts. AI neither helps nor harms here, it just makes implementing good and bad decisions faster.
It's like if I asked an LLM to produce a contract for a a new hire
(my friend the one I mentioned, did that BTW LOL)
Yeah, okay, it might be good, probably very good, in fact might be GREAT.
But I just have no clue
If I am understanding you correctly, you are claiming that an AI will not design a table to be readable unless a programmer with good taste is prompting it. Did I understand you correctly?
If so, then yes we are going to arrive at different conclusions, because my claim is that the table made by the programmer with no taste will be acceptable to the client, hence good taste won't save the other programmer because good taste will require more guiding and interference by the programmer. And then the client will just go with the no taste implementation.
If the programmer with good taste truly does not lose any time implementing their solution when compared to the other programmer, as you claim, then yes the client will go with good taste since all else are equal.
I just find those assumptions hard to believe at this stage.
1. I did consulting for a bit, $500/hr haha I was so reluctant to do it. I do admit that SV was just pouring way too much money back then. The good old days. I don't live in the US, and remote opportunities have been vanishing quite quickly. In the US I would have no doubt I'd get crazy good offers, I have lots of people who respect me there. My friend and I even got offered high positions at Youtube way back when.
2. I quit 3 years ago and have done a lot of artisanal code. Lots of fun. But next year I'll need to figure out stuff, hence why I accepted this small contract (ends next month) and thinking if it's time to move on.
BTW I have a family to feed and my son eats like a husky even though he's only 45 pounds, so there's that. My daughter is a big fan of Argentinean shrimp and caviar. You get the point.
Hitting you an email
"C’est une folie de haïr toutes les roses parce qu’une épine vous a piqué."
I already lived under a cave but did you see the amount of money and hype pushed into AI? Just look at YC, 90% AI, just like it was crypto back then.
Jensen Huang saying we have AGI.
One thorn I can handle. A field full of slugs and nettles with billions of dollars to burn I'm not so sure
But of course I am teaching CS, and now it's just a nightmare of trying to figure out how to teach it. We have a lot of people saying we still need to teach the basics in addition to AI-assisted, and we have a lot of people saying to just teach vibe coding, and a lot of people saying just teach SE and no coding. And we're just trying to do the best we can for the students, while wondering if the field even has a sensible future.
(Anecdata: universally, my hired grads have told me they are incredibly grateful to have gotten through college before LLMs hit the scene.)
And it's a pisser, because teaching CS was great. I get to try to optimize learning and talk about what I love? Heaven.
I think I'm in this category of software engineer in terms of how much I love the work:
Requirements gathering: 5%
Functional spec: 5%
Design doc: 20%
Programming: 65%
Deployment: 5%
And engineers who were here:
Requirements gathering: 20%
Functional spec: 25%
Design doc: 25%
Programming: 5% (or 0%!)
Deployment: 25%
are having a much easier time with this transition. Unfortunately for me, programming is also the most fun to teach.
I have a lot of other interests outside of CS, one of which would actually pay money (GIS). Not sure how to get paid for dualsporting, hiking, bicycling, geology, visiting museums, writing (I do get paid grocery money for this), hand-coding, or reading. But I'm open to suggestion. :)
I also love teaching (not particularly great at it, but not the worst) and taught SALSA dancing for a few years. The money was okayish (enough for myself but not to feed my children) but I LOVED being around ppl.
I wasn't talented in dancing but my talent in programming really made me happy (and well paid), so that kind of sucks. Definitely same as you, I'd even say programming can be my 90% of joy and effort (in a bad way sometimes, but I manage to talk myself into dialing it down)
I can't imagine the hell teachers are going through, especially coding, with AI.
Luckily I have a lot of weird interests and some can pay well (for example, I've had a decent success teaching people how to lose weight doing intermittent fasting, imagine the randomness of my interests)
What is your next plan?
I wasn't joking on becoming a yoga teacher, as cliché as it sounds, because I did it and quite successfully so (I was young, handsome, and single back then, and my classes were PACKED. I'm still handsome but the other two are gone. People paid cash so I had a lot of bills in my home lol)
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'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.
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
Unfortunately, there is demand for people who know to sanity check things. It's not glamorous, but neither is damage control.
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.
We're not a company, it's just me and my pal. I keep 50% of the product and profits and he pays me to do that (because I have way more expertise than him)
curious question where you work, the AI craze is not happening?
of course it is. There's no escape IMO unless, as you've said, go solo or with a small team. I just don't know of any company of a notable size that's not into it full force. It's both disgusting and amazing to look at really. Give it a few years and if it fizzles out I'm curious what will happen. If it turns great, I'm even more curious what will happen. Weird times.
Just "yesterday" we were all careful what and how we do stuff, you have now people who barely understand things slopping together OS and people use it blindly (Omarchy), and then shocked pikachu faces when giant gaping holes are noticed in it.. and people still do not care. Amazing trainwreck of a timeline, hah.
I'm seriously considering pivoting myself to security.
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.
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.
As much as I like to put power in the individual, this has become a huge nightmare for those of us who were kind of decent programming.
This was NOT the case 5 years ago. At all.
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?
The problem is these companies literally say "NO NEED TO LEARN ANYTHING JUST VIBE". Ton of ads on social media and YT claiming that.
Like my mom doesn't even know what a PC, a Macbook or a Linux computer is and we're telling her she can make apps?
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.
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.
This is all but gone
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.
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.
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.
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.
That's pretty much my take on it, yeah.
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.
A simple example: with picture editing software people have been able to create believable fake content for a long time, it did require skill which limited the scale, with genAI creating in seconds, by anyone, a much more plausible fake image than most people were able to before creates a whole other kind of problem that didn't exist before.
It's a tired argument to reduce issues to analogous which existed before and discard how the scale and degree of it changes the class of problem...
> A simple example: with picture editing software
Actually reminds me of a story a professor in a comsci ethics class once told about how the police in South Africa (during the apartheid years) would take photos of everyone protesting the regime at the university campus. Back then (this was the 80s) nobody really cared - what were they going to do, go through all the photos by hand?
He said "imagine how much more concerned people would be today". That was in 2011 or so.
Well that didn't last long. The temptation was too big for him too
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
so, would you say "no worries man I know what I'm doing" about how you use AI?
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.
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.
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.
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.
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.
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.
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.
My guy in fact started by blaming himself for prompting wrong, and yesterday he was "upset" that his friend Claudio missed some important documentation details (three API keys were needed and one I had no idea where to get, not mentioned once at all)
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".
and yeah luckily I ended up quitting just right when coding was getting easy with LLMs and even back then (3 years ago) I remember our PMs and bosses saying "look why does it take you a week? I did in in two hours!"
Got riddance to get out of those jobs, BUT I do agree that paying customers != shipping software. Getting paid customers is a great skill and it's the skill my friend has (he's an A class marketer, and not in a fishy or scammy way). The guy just understands marketting. That's why I'm working with him.
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
Sucks that my children have to smell it, really different from my chilhood.
I don't know almost anyone who preferred having a childhood today (with so much social media and doomscrolling) as compared to the 80's and 90's, when it was just kids playing hide and seek outdoor, a few Michael Jackson home-recorded cassettes, and of course Sonic 2.
The FE team consists of me btw.
And for those who say "well why code and use your brain?", may I remind them that cities where people do a lot of walking are so much enjoyable, better for people mentally and physically than places where people use the "efficient and smart way" of transportation, that being a car.
I remember my cousins in the States picking me up by car and walking their dog by car, inside a gated community that was very enjoyable to walk. Pathetic to see.
(BTW both my cousins are severly overweight, 200+ pounds)
This was not feasible a year ago.
Now, regarding the quality of the code, the edge cases etc, surely someone familiar with the failure modes can review and prompt the AI better. But that doesn't negate the fact that these people are producing huge amount of code using natural language without any programming background.
A skill that used to take many months (if not years to master) and require handcrafting the software line by line is now done in English with almost no understanding of programming primitives.
And I'd argue that if the models, harnesses and IDEs keep getting better then a layman learning few software best practices would outperform a team of senior developers writing code manually.
Therefore, the developers refusing to adapt and adjust will find themselves gradually in a position where a layman can easily outperform their economic output by a huge margins. And in my opinion, the author just found himself in this position, and still in denial of what is happening.
Huge amount of *non-working* code, as illustrated by this article.
Also author is not super clear on this point, but according to article the partner has programming background.