I'm at the opposite end. I feel AI is sucking all the joy out of the profession. Might pivot away and perhaps live a simpler life. Only problem is that I really need the paycheck :(
They've even got their own slogan: "you're probably just not prompting it properly"
That's the same energy as telling other professions to "just learn to code, bro" once they are displaced by AI.
But I guess it doesn't feel nice once the shoe is on the other foot, though. If nobody values the quality of human art, why should anybody value the quality of human code?
It's the exact same neoliberal elites who told everyone to code one year and told them they'd all be automated of a job the next year.
I dunno who exactly you think you're being condescending towards.
Also we are still designing systems and have to be able to define the problem properly, at least in my company when we look at the velocity in delivering projects it is barely up since AI because the bottlenecks are elsewhere..
Do you truly believe it won't get better, maybe even better at whole system design and implementation than people?
Literally yesterday I remarked to my tech friends how fun coding with CoPilot is. I actually make forward progress now, and I understand all that the agent is doing.
For me, coding is an enjoyable means to an end. I do enjoy the process, but I enjoy the results more.
You're right though about it not choosing some different path, I might or I might not know that.
What are you calling "growth"? Adoption, or LLM progress? LLM progress has objectively slowed down, and for rather obvious reasons. The leaps from GPT-2 to GPT-4 can't be reprised forever.
I don't think it will be; a vibe coder using Gas Town will easily spit out 300k LoC for a MVP TODO application. Can you imagine what it will spit out for anything non-trivial?
How do you even begin to approach remedying that? The only recourse for humans is to offer to rebuild it all using the existing features as a functional spec.
There are cases where that will be the appropriate decision. That may not be every case, but it'll be enough cases that there's money to be made.
There will be other cases where just untangling the clusterfuck and coming up with any sense of direction at all, to be implemented however, will be the key deliverable.
I have had several projects that look like this already in the VoIP world, and it's been very gainful. However, my industry probably does not compare fairly to the common denominator of CRUD apps in common tech stacks; some of it is specialised enough that the LLMs drop to GPT-2 type levels of utility (and hallucination! -- that's been particularly lucrative).
Anyway, the problem to be solved in vibe coding remediation often has little to do with the code itself, which we can all agree can be generated in essentially infinite amounts at a pace that is, for all intents and purposes, almost instantaneous. If you are in need vibe coding disaster remediation consulting, it's not because you need to refactor 300,000 lines of slop real quick. That's not going to happen.
The general business problem to be solved is how to make this consumable to the business as a whole, which still moves at the speed of human. I am fond of a metaphor I heard somewhere: you can't just plug a firehose into your house's plumbing and expect a fire hydrant's worth of water pressure out of your kitchen faucet.
In the same way, removing the barriers to writing 300,000 lines isn't the same as removing the barriers to operationalising, adopting and owning 300,000 lines in a way that can be a realistic input into a real-world product or service. I'm not talking about the really airy-fairy appeals to maintainability or reliability one sometimes hears (although, those are very real concerns), but rather, how to get one's arms around the 300,000 lines from a product direction perspective, except by prompting one's way into even more slop.
I think that's where the challenges will be, and if you understand that challenge, especially in industry- and domain-specific ways (always critical for moats), I think there's a brisk livelihood to be made here in the foreseeable future. I make a living from adding deep specialist knowledge to projects executed by people who have no idea what they're doing, and LLMs haven't materially altered that reality in any way. Giving people who have no idea what they're doing a way to express that cluelessness in tremendous amounts of code, quickly, doesn't really solve the problem, although it certainly alters the texture of the problem.
Lastly, it's probably not a great time to be a very middling pure CRUD web app developer. However, has it ever been, outside of SV and certain very select, fortunate corners of the economy? The lack of moat around it was a problem long before LLMs. I, for example, can't imagine making a comfortable living in it outside of SV engineer inflation; it just doesn't pay remotely enough in most other places. Like everything else worth doing, deep specialisation is valuable and, to some extent, insulating. Underappreciated specialist personalities will certainly see a return in a flight-to-quality environment.
> If you are in need vibe coding disaster remediation consulting, it's not because you need to refactor 300,000 lines of slop real quick. That's not going to happen.
My experience as a consultant to business is that they only ever bring in consultants when they need a fix and are in a hurry. No client of mine ever phoned me up to say "Hey, there, have you any timeslots next week to advise on the best way to do $FOO?", it's always "Hey there, we need to get out an urgent fix to this crashing/broken system/process - can we chat during your next free slot?".
> Like everything else worth doing, deep specialisation is valuable and, to some extent, insulating.
I dunno about this - depends on the specialisation.
They want a deep specialist in K8? Sure, they'll hire a consultant. Someone very specialist in React? They'll hire a consultant. C++ experts? Consultants again.
Someone with deep knowledge of the insurance industry? Nope - they'll look for a f/timer. Someone with deep knowledge of payment processing? No consultant, they'll get a f/timer.
No, that's fair, and I think you're right about that. But refactoring 300,000 lines 'real quick' isn't going to happen, regardless of that. :)
> They want a deep specialist in K8? Sure, they'll hire a consultant. Someone very specialist in React? They'll hire a consultant. C++ experts? Consultants again.
I implicitly had narrow technical specialisations in mind, albeit including ones that intersect with things like "insurance industry workflows".
Like 80% of jobs outside the USA are either local or outsourced CRUD web applications. Many people live quite well thanks to exchange rates. I wonder what's gonna happen if/when those jobs disappear.
Businesses don't pay for CRUD apps, businesses pay for apps that solve problems which often involves CRUD to persist their valuable data. This is often within the sometimes very strange and difficult to understand business logic which varies greatly from one business to another. That is what "CRUD app developers" actually do, so dismissing them as though there is zero business logic and only CRUD is doing them, us, a disservice.
Why, I do plenty of what you describe myself...
That's my worry. Might be put off a few years, but still...
This is a possibility in very well-trodden areas of tech, where the stack and the application are both banal to the point of being infinitely well-represented in the training.
As far as anything with any kind of moat whatsoever? Here, I'm not too concerned.
For what I am vibing my normal work process is: build a feature until it works, have decent test coverage, then ask Claude to offer a code critique and propose refactoring ideas. I'd review them and decide which to implement. It is token-heavy but produces good, elegant codebases at scales I am working on for my side projects. I do this for every feature that is completed, and have it maintain design docs that document the software architecture choices made so far. It largely ignores them when vibing very interactively on a new feature, but it does help with the regular refactoring.
In my experience, it doubles the token costs per feature but otherwise it works fine.
I have been programming since I was 7 - 40 years ago. Across all tech stacks, from barebones assembly through enterprise architecture for a large enterprise. I thought I was a decent good coder, programmer and architect. Now, I find the code Claude/Opus 4.5 generates for me to be in general of higher quality then anything I ever made myself.
Mainly because it does things I'd be too tired to do, or never bother because why expand energy on refactoring for something that is perfectly working and not to be further developed.
Btw, its a good teaching tool. Load a codebase or build one, and then have it describe the current software architecture, propose changes and explain their impact and so on.
I have about the same experience as you do and experience using Opus 4.5.
If this is true, you weren’t a very good programmer. There’s much more to code quality than refactoring working code.
Yup, my conclusion exactly.
With that said, most code I have seen in private sector is almost objectively horrible (and certainly subjectively). Code manufactured with the current best tools such as Claude compares favourably. Companies rarely have the patience to pay for well manicured, elegant code. If it sort of works it ships.
A good engineer will tell you how to spend 25% of effort to get to 90% of the result you want. With maintainable code, and importantly with less code that touches fewer systems.
A bad engineer will deliver exactly what product asked for without asking questions, generate 4x the code, and touch every piece of the system.
Companies are just setup in a way that incentivizes building organizations that create bad code. Most places would rather hire 100 bad engineers who can be easily replaced than 5 good engineers.
This is quite true, and it is this -- really, a special case of "the market can remain irrational longer than you can stay solvent" -- that has me worried about the implications for the labour economy more than anything else.
Despite having worked with or managed hundreds of developers over the years, I don't think there were more than, maybe, 5 that I would not gladly swap for a claude-equivalent. Diligent, able to produce good code when adequately supervised, and devoid of a desire for work life balance :-)
Most software devs are just mediocre, they learn to code late in life, they work 9-5, and often stop learning as quickly as they can.
The way the do, which is? I've skimmed comments and a lot of them is hate, hostility towards OP's project and coders "without skill" in general, also denial because there's no way anything vibe-coded worked. At best, there is strong tribalism on both ends.
You might see more opposing views in this thread, but if you browse this site often you'll see both sides.
Those embracing it heavily do not see the nuances carefully creating maintainable solutions, planning and recognizing tech debt, and where it's acceptable short term. They are also missing the theory building behind what is being created. Sure AI models might get even better and could solve everything. But I think it's naive to think that will be generally good for 90% of the population including people not in tech.
Using these models (text or image) devalues the work of everyone in more than one way. It is harmful for creative work and human expression.
This tech, and a lot of tech, especially ones built by large corporations for profit extraction and human exploitation, is very unlikely to improve the lives at a population level long term. It can be said for a lot of tech (ie. social media = powerful propaganda). The goal of the people creating these models are to not need humans for their work. At which point I don't know what would happen, kill the peasants?
I agree what we do requires a lot of creative thinking. When AI supporters attempt to use an argument comparing to factory workers being freed from dull laborious work by robots, the analogy falls flat on two fronts. First, there's nothing creative about that sort of work and second, because robots are highly accurate; while AI can often be just high.
Just like SEO experts, marketing experts, trade bots and crypto experts; the vibe coders will weed out.
is there a term for that?
AI at our fingertips, accessible and useful, that's just a tool, that's not redefining us as an industry and denying people's jobs – that's an asset. (I used an em dash to prove I am not AI, as apparently double dash is now a sign of AI text!)*
(*) case in point, the situation is _TIRING_.
It's a miracle. Simply wouldn't have been done before. I think we'll see an explosion of software in small and midsize companies.
I admit it may be crappy software, but as long as the scope is small - who cares? It certainly is better than the janky manual paper processes, excel sheets, or just stuff in someone's head!
1. Invoice billing review. Automated 80% of what was a manual process by providing AI suggestions in an automated way. Saved 3 hours per day of managers time. Increased topline by 10%. Dev time: 1 day
2. Data dashboards. We use janky saas that does not have APIs. Automated a scraper to login, download the reports daily, parse and upload to a database, and build a dashboard. Used to take my associate 3 hours per week to do this in a crappy spreadsheet. Now I have it in a perfect database much more frequently. Dev time: 4 hours.
We are attacking little problems all across the business now.
A MIRACLE!!!!
I wouldn't want to hassle customers who have fully paid up accounts
I think also you need to compare it to what was already there. No QA on the humans. Done off the side of their desk with no oversite, process, or checking. Huge amounts of manual errors.
The new solution just needs to be better than the old one, it doesn't need to be perfect.
(But I 100% agree that I wouldn't let AI live against customers. It is helping us build automations faster, and doing a "little" thinking on recommendation rules that would be very hard to implement without something highly structured, which would be frankly impossible in our environment.)
No. The bar is "miracle" and can cure cancer etc and can replace all developers etc. The bar is much higher than existing manual processes. It absolutely needs to be perfection to match the lofty claims
I guess Vibe coding cleanup firms and offensive security researchers are plotting to find bugs costing firms millions of dollars worth of bugs or one creating a dreadful data breach.
And to be honest, even the tiny apps I'm doing I wouldn't have been able to do without some background in how frontend / backend should work, what a relational database is, etc. (I was an unskilled technical PM in the dotcom boom in the 2000s so at least know my way around a database a little. I know what these parts of tech CAN do, but I didn't have the skills to make them do it myself.)
Funnily enough, Excel is the quintessential example of a fourth generation language, IDE, and database and it's the only one aside from SQL which actually succeeded from its time period. It's software, just like what you're building now, and just like what you're building now there are good points and bad points about it. The tradeoffs are different between the JS / Python code you're likely spinning up now vs. the Excel code that was being spun up before, but they rhyme.
For me, that is nightmare fuel. We already have too much software! And it's all one framework or host app version update away from failure.
Most developers are too full of themselves, in fact, most of us are a bunch of pretentious pricks. It is no wonder people are happy to be able to get what they want without our smugness and pretentiousness. Too bad some us are not like that and will end up getting unemployed anyway in the next few years.
"Comment NEAT to receive the link, and don't forget to connect so I can email you" -- this is the most infuriating line ever.
Im in this field and my system was heavily built with Claude, though not per vibe coding, more like a junior supporting me: I do not see any person connecting a vibe coded bot to a real account soon, since if its about real money, people will hesitate. And if you have blown up one account with your vibe coded bot while you are not a professional dev, you will loose interest very quickly - such systems do not contain "just a few thousand lines of code": Sure you could speed up development massivly and "hit the rock sooner than later" when going vibe coded here :-D
In comms, they have something like a 1:4 ratio of design to validation engineers. Defence is slightly different, as it depends on the company, but generally the tolerance for bugs is zero. Lets not get started on the HF trading folks and their risk appetite!
There's a lot of room for software engineers. Most FPGAs are SoC devices now, running some form of embedded linux doing high-level task management networking. Provided you know enough Verilog to know your way around, you'll be fine. You're also in a space where most engineers I know are preparing to retire in the next 5-10 years, so there will be a panic which will ripple across industries.
FPGA basics: https://nandland.com/fpga-101/
Verilog basics: https://hdlbits.01xz.net/wiki/Main_Page
Projects: https://www.hackster.io/fpga/projects
There are people here "I can finally get all my ideas done!" Sure, if they are really important enough, I guess. But high technology is much, much less important to me than my employer or probably others here on HN. I can only be concerned with the paycheck at this point. And at this point, they are happy that I can read documentation, write code, read documentation, write code, and don't care how it gets done. (For what I am working in though, I'd just skip the AI training step.)
With that in mind, I like to use PLs as tools to clarify thinking. There are others that think using PLs and their accompanying tools as friction to their goals, but my friction is understanding the problems I am trying to solve. So, while taking the adventure into automated tooling might be interesting, it doesn't replace the friction (just the feeling I have to read more potential garbage code.)
It's also a depressing wakeup call to realize that programming has evolved from a craft in which you used to write 90% of the instructions but with the rise of libraries, and now codebots, 99% of the instructions are written by others. Coding became cut-and-paste decades ago but now it's degenerated into talk-and-walk. Soon there'll be no need for any skill from the code creator at all. The writing is on the wall. Frankensteinian LLMs surely will drive all the engineers from the building.
It was great while it lasted, but... sayonara hackerdom.
Security engineers will have jobs until software is perfectly secure... and that is going to be a while.
I do not use LLMs at all to do my job, and it is unlikely I ever would. Clients pay me -after- they had all their favorite LLMs take a pass.
Might be never or if the software is not used at all.
The perfect and secure software is none.
Well, at least not connected to the internet?
Most of my job as a security engineer is making offline systems you can actually trust.
And indeed the vibe coders will just create a lot more security issues
Not as long as you think.
https://cybernews.com/security/standord-artemis-system-beats...
I think the coding tools are not good enough yet so we can kinda-sorta hang on, but they will be within a few years.
These existed before but the culture surrounding AI delivered a double dose of both.
I have no problems with LLMs themselves or even how they are used but it has developed its own religion filled with dogma, faith based reasoning and priests which is utterly toxic.
The tools are shoved down our throats (thanks to the priesthood, AI use is now a job performance criteria) and when they fail we are not met with curiosity and a desire to understand but with hostility and gaslighting.
I am of the same age. I have some good ideas on where to go, but dread the grind to get things moving. When I was in my teens and 20s the grind that got me to where am now was fun, but doing it again looks far less appealing now.
Marvel Champions in particular is a lot of fun, although may be a bit overwhelming at first if you don't play a lot of board games already.
I also got into Legendary deckbuilding games recently, and those are a bit more approachable, although not all of them play solo unless you manage two hands of cards (which isn't a big deal for me, but I've played hundreds of different board games).
They have those based on various IPs (Game of Thrones, James Bond, X-Files, Matrix, Alien movies, Buffy, Marvel, and in a few months DC comics) and play somewhat similarly, so if you learn one it would be easy to learn another one.
I also picked up a solitaire variant called Hoki just last week and really enjoyed it. You upgrade your cards over multiple games (that are each about five minutes to play), and then once you've completely upgraded all the cards you can play the game daily and then consult a book that will give you a fortune based on the final state of your game.
It took me 53 games to unlock the final state, and I did all of them in just a couple of days, I enjoyed it so much. Now I'm playing a game or two a day to see what the fortune is and then writing a journal to reflect on what that could mean, for fun.
Slowly getting back into my creative hobbies this year (which include board game design and writing), although coding I still feel is hard to do in my off time (even when it's making games, which I've historically really enjoyed doing).
I've messed around with A.I. agent coding a bit, and I'm a bit more impressed with it than I anticipated, but I'm not sure how deep down that rabbit hole I want to go and not code myself. But I really don't feel like I have much energy left in the tank for coding more after doing it for my day job lately.
But ya, I was out of work for 9 months and enjoyed pretty much every minute of it up until I realized it was not as easy to find a job as I thought it would be (well, it played out a bit differently but for the sake of brevity I'll stick that story)
A lot of work was tedious, painstaking grind, but the reward at the end was considerable.
AI has completely annihilated all of the joy I got out of the process, and everything that attracted me to it with such abandon as an adolescent and a teenager. If someone had told me it was mostly slop curation, I would have stayed in school, stuck to my philosophy major, and who knows -- anything but this. I'm sure I'd have got reasonably far in law, too, despite the unpropitious time to be a JD.
I'm still working on my own small closed source projects, building them the way I want to, like a gameboy emulator - and I've gotten a lot of joy from those.
No matter how 'senior' you are, when you lose touch with the code, you will, slowly, lose the ability to audit what LLMs spit out, while the world moves on. You got the ability to do that by banging your head against code the hard, "pre-AI" way, perhaps for decades, and if you don't do the reps, the muscle will atrophy. People who think this doesn't matter anymore, and you can just forget the code and "embrace exponentials" or whatever, are smoking the good crack; it _is_ about the code, which is exactly why LLMs' ability to write it is the object of such close examination and contestation.
Folks who realise this will show to advantage in the longer run. I don't mean that one shouldn't use LLMs as an accelerant -- that ship has sailed, I think. However, there is a really good case to be made for writing a lot by hand.
One day I might start a consultancy business that only does artisanal code. You can hire me and my future apprentices to replace AI code with handcrafted code. I will use my company to teach the younger generation how to write code without AI tooling.
That's an interesting perspective. I guess it depends on what you want and how low the stakes are. Artisanal coffee, sure. Artisanal clothing, why not? Would you want an artisanal MRI machine? Not sure. I wouldn't really want it "hand crafted", I just want it to do it's job.