Just to make this clear up front: I was always very skeptical of LLMs (I’ll never call them AI) and still see them as fundamentally limited and unreliable. See my post history if you don’t believe me.
Having said that: I work in business software and the models have progressed to the point where typing code and writing classes and functions by hand has become uneconomical. We review the output closely to avoid the idiocy of “vibe-coding”, but the quality has improved to the point where we rarely have to outright reject solutions. The discussions have moved to high-level architectural points, where we just request better implementations than the first-shot. We get them, review again and move on. Maybe there are deeper, latent problems beeing embedded in the code-base through this process, but we can’t spot any although we are looking intently.
And just like that, my profession has died. Nobody in my team writes code anymore. Things that would have taken days are now done in minutes. We are all just considering design and architecture. We are changing our processes since they are not fit for the new way of doing things. Scrum? What’s even the point anymore? We can work through a sprint-backlog that would’ve taken two weeks in two days now. We do reviews 90% of the time. When we request changes they come back in minutes.
I have a feeling like the ground has been taken from below me. My son wanted to become a developer, that isn’t a real job anymore. Everyone around me feels the same way, including our management. We are a small team and a small company, everyone is effective and our jobs are not in danger as long as our product is succesful. But my profession is essentially dead. It’s a craft I honed for many, many years and the craft itself is now automated. I wanted to improve my typing skills to code faster, this is now a ridiculous endeavour. We don’t know what comes next, but I now only have my analytical skills and knowledge of good architecture to fall back on.
And it all changed so fast. I don’t think this will transfer to other professions so easily, but for programming LLMs ended it all in the span of the last four months. I thought I knew what my future holds as a senior dev, now I can barely see past christmas.This is a nightmare.
I’m in the exact same boat. I’ve been thinking a lot about how it’s changed the landscape of tech, and ive come to an interesting conclusion so far. It feels kind of like the switch from horse and buggy to motorized vehicles.
But, what’s going to happen after these companies bankrupt? Will AWS still host old Anthropic models at an increased price? It feels like when Walmart moved into small towns and sold their products at a loss in order to kill mom and pop stores.
Massive enterprises will cut deals to keep using frontier models, everyone else will have to go back to using their atrophied brains. Those who have the compute will be running local models.
Similar to how employers today ask if you have a vehicle to get to work, they will ask if you have local inference. It hurt, but, I bought an AMD Halo the other day. They turned the information super highway into a stroad and the Halo is my vehicle.
This is the exact same thing that happened with luddites, which today we know as ‘idiots who oppose technological progress’ thanks to capitalist propaganda.
It was never about efficiency, quality, quantity, or accessibility. It’s almost exclusively about getting people to stop doing things for themselves, and instead rely on expensive machinery owned exclusively by the capitalist class for everything.
Most average people used to make and fix their own clothing, which meant they had little need for corporations that produce the same clothing, but lower quality. So capitalists forced everyone to use machines, until nobody remembered how to make clothes without them.
Same with software, languages, and a lot of other skills and office jobs right now. Educated people who know how to do stuff are not only unprofitable, but also dangerous. Make them all use AI until they forget everything and have to rely on megacorporations for the tiniest thing.
This is all the more reason to keep reading books, learning, DIY as much as possible, coding your own software, and even learning stuff like woodworking, sewing, farming, etc.
Being educated, informed, knowledgeable, and skilled is among the easiest, but at the same time most effective acts of resistance.
Do not let them turn you into mindless slaves.
Well, just today I had 3 meetings with different AWS techs about why a policy wasn’t working or why users couldnt use certain services. Turns out all of them used AI to generate the aws policy they tried to use and the policies were wrong. The techs were rookies and lacked the understanding needed to correct the code. I was able to fix each and rewrite policies by hand.
Now when I and the rest of my greybeard generation start to retire, the rookies coming in will not have the exposure to code, policies, etc that we had. The knowledge gained from trial and error will disappear with the retirees. That’s when the crisis will really begin. When we in the know will not be around to correct the rookie mistakes.
Here’s the thing, though. Give it a generation (and in this industry, a “generation” is basically the amount of time it takes one class of grads to get through college, or possibly at the outside maybe two) and there will be no one left with the skillset required to effectively babysit the LLMs anymore. Everyone who used to have those skills but leaned on AI and stopped coding themselves will have let them atrophy, and the upcoming crop of junior developers won’t know how to do anything but prompt the chatbot.
Maybe there are deeper, latent problems beeing embedded in the code-base through this process, but we can’t spot any although we are looking intently.
It’s this right here. Nobody will be able to look intently, because no one will have the knowledge of what to look for anymore. You are able to effectively manage the stupid machine’s output because you actually know more or less what functional code looks like and how it works. Once that skill has been squeezed out of the industry it’s going to be utter chaos.
Yes, that’s pretty obvious to anyone in the industry. But like many catastrophies mankind can see decades ahead, nothing will be done about until the consequence bite hard. I’ll be old by then.
I’m sort of surprised there’s no conscious effort to counter that. In other areas where automation has taken over they will ensure that people can still do the work manually. I visited an automated metro system a while ago, and they explained that even though certain lines are GoA 4 (in other words so highly automated they don’t need staff on board anymore), they still have people operating the trains manually every once in a while, specifically so that they don’t forget how to do it if need be, and that the new recruits can learn as well.
I’m sort of surprised there’s no conscious effort to counter that.
Nobody wants to be the company “wasting” money on training young coders. This is the inevitable decay caused by Capitalism.
That probably wasn’t a privately run Metro then. In private industry it’s usually cheaper to risk an accident than pay for all those man hours.
Yes, to be fair it was, at least practically. It was one of these government owned companies - so that they’re still expected to work kind of as a private company. So sort of an in between thing.
Are you? I’ve not seen many people ever pay much attention to potential problems until it’s way too late.
It’s like how you can walk into a restaurant and retail establishment now and it’s not uncommon to interact with an employee that doesn’t know how to do the simple math required to make change during a transaction.
I’m stuck using MS Teams at my job, and I’ve noticed changes to the UI, removing of visual cues and triggers, which I can only assume they’re doing to get people to ask their stupid chatbot how to make a change.
Eventually this is going to bite hard, but it’ll only bite the working class, so the people in power do not give a shit.
Almost a graduated here, here’s my small rant (kinda).
For context: I rarely use LLMs cuz there’s no fun on vibe coding, not to mention the shit ton of problems its usage carries (economical, environmental, social, ethical, skills decline, etc)
The other day I was in the classroom using my Actual Intelligence (AI) to do a python project, when a group of randoms came by and asked what I was doing, they were looking at me like if a was some sort of caveman, because apparently I’m one of the few guys that still codes by hand.
Sometimes I feel bad for them, it’s depressing to see everyone getting stupider so quickly, but I can relate to that feeling of wanting to do absolutely nothing from day to day.
In most of my classes, I’m the only one answering questions to the point that sometimes I get banned from answering (although I don’t know if it’s because everyone is kinda shy or if it’s truly because they don’t have an idea of what’s being talked about)
My own experience of almost 2 decades as a Freelance senior software developer, mostly coming in to work on top of and even fixing existing codebases of large custom-made applications which were for years in Production in mainly large companies, which generally had layer on top of layer of different coding styles and even design approaches because multiple people worked in them each thinking “I know best” and doing their own thing instead to trying to work within the existing style, is that getting to the point of understanding it at all levels takes many months, and that’s even after one is experienced enough at it to have discovered shortcuts like learning and identifying the coding and design style of specific engineers and even their typical mistakes (people tend to be pretty consistent in the mistakes they make) as a way to accelerate understanding, detection of problems and figuring out were and how it’s safe to change code and design without it cascading to problems elsewhere.
All this to say that picking up and fixing a real life system in Production which was mostly AI created and hence doesn’t even have the patterns for coding and design choices and mistakes that humans have, will take many months, maybe more than a year PER SYSTEM even for somebody who is senior and highly experience at it.
So fixing the AI produced stuff will have giant costs which probably more than offset the savings at creation time from using AI.
Now, some would say that you can just get the AI to regenerate the whole thing to fix it, however that has two problems:
- You’re just replacing old known bugs with new and entirelly different bugs. Also any UI will change. Good luck selling such changes to the system they use daily to the business side.
- In my 3 decades of experience in the Industry, clear, complete, detailed and unambigous Requirements and Interface Definition documents are as rare as hen’s teeth and without such things you can’t prompt the AI to make software that yields the same user functionality and talks with other systems correctly. So features will change or behave differently, if it’s good enough integration testing will fail in brand new and exciting ways, if it’s not good enough it will blow up in Production or, even more entertainingly, Production databases will be filled with broken data.
So, yeah, full regen just at best moves the problems around and pisses users off.
It’s this right here. Nobody will be able to look intently, because no one will have the knowledge of what to look for anymore. You are able to effectively manage the stupid machine’s output because you actually know more or less what functional code looks like and how it works. Once that skill has been squeezed out of the industry it’s going to be utter chaos.
I see no reason why people can’t learn to do this task.
I learned to write an operating system kernel in college. Not because I would ever need to do that in a job, but because it was considered important to learn how stuff worked, to gain a deeper understanding to make architectural decisions. I also learned to write assembly, for the same reason. Similarly, good educations in the future will likely teach people to code, even if they will never code in their jobs.
That’s like saying few are left knowing assembly. It’s just not needed anymore. Could it be needed? Sure. Are you going to get a job easily with that? No.
It’s just not needed anymore.
Someone needs to maintain compilers though right ?
One child policy has done a number on China. Even more sinister example is Ethyl gas or gasoline with lead. that went for almost a century. Was also just wondering what ancient cities were like when they dumped their waste out on the street.
It is absolutely astounding our species has survived as long as it has.
Why would new learners not learn what to check…?
When we went from machine code to assembly people were very negative about it. Same again when moving to high level languages. And of course this continued with every single development since then and now it continues with LLMs. Always negative about it, no more skill, etc. the exact same arguments really.
This argument is flawed, as I just took a break from reading disassembly to scroll a bit. Assembly is the human readable representation of machine code. It is relevant today (eg embedded development).
Also each move to higher level comes with its own set of issues and trade offs. I think you somewhat correct to correlate with the past advancements, but LLMs fundamentally different: It is not a new higher language, its normal language. Also it cannot grab an oscilloscope and debug HW issues which are wrongly documented by the supplier.
Keep holding on. LLMs are still in the operate-at-a-loss-to-control-the-market stage. Once they start charging the real cost, human software engineers will be more in demand. That and refocus. All of human history is careers reinventing themselves as new tools become available.
Here is a new mission for all the Post-LLM-Depression existential crises developers:
Community Computing.
Use your skills to build a network with big tech alternatives for your community! Federate your network with other likeminded people’s networks. Restart the internet with real people without bots, ai agents, or advertisers.
Don’t quit your passion! Use it as fuel to motivate yourself and help others move away from big tech!
The hard part is getting paid to do it.
That’s a volunteering organisation, no?
It looks like a non profit structure with membership dues
Maybe you aren’t talking about this but I think we software folks have to get used to the idea that the big huge paycheques are not coming back our way.
You don’t get paid that kind of money in this economy unless someone is getting screwed over.
Man I love this comment so much. I’ve dreamed about doing this kind of thing
In my experience that’s only been the case for functionally non-unique software.
At my job we write software that’s effectively unique. There isn’t a pattern to fall back on.
It makes a lot more mistakes very frequently, since the nature of it keeps trying to guide it to the known path. It’s great for debugging since the manner that some check failed is chasing through known logic, and it doesn’t matter that it doesn’t see the existing code as probable or fitting a known pattern if the shape of a switch statement is still the same.In my personal projects I’ve even run into issues getting it to do stuff like vulkan synchronization barriers in a way that doesn’t continuously devolve into what you see in tutorials, with an ultimately unstructured pile of adhoc resource allocation and barriers.
It will cause a massive loss of human knowledge, nobody would learn how because it’s just a prompt
And it’s not only for coding. Yesterday I was shown that because they probably stole a lot of parametric designs Claude can even do 3d files from photos and quotes, it spits out an openscad file and compiles it. I was shocked, it added chamfering and small details, from a rough sketch photo. Who’s going to learn how to use a cad program if it can be done in 3 minutes from a prompt? And the process of learning taught you all the quirks , but in the future who’s going to be a pro?
Tbh I legitimately do not buy this. I use the best models, top of the line harness and the code still sucks it sucks less granted, but it’s still just not good lol.
If I’m ever in the situation that “typing the code” was a bottle neck instead of figuring out what to type then I know I’ve fucked up somewhere.
The worrying part for me is how the fuck do we keep people knowing good code and architecture and able to review the output without practicing the art for a couple — or three — decades, first? It just won’t happen, and then the new generations of “devs” won’t have done any coding, basically, and they won’t have learned anything the hard way. Security issues are going to be a nightmare in the future, no?
This is where Idiocracy becomes real. People won’t know how to do the things to make sure they’re done right. I’d say it’s like history where jobs that used to exist no longer do because they were replaced by something better, faster, and cheaper…but those were individual jobs and not what LLMs are being shoehorned into essentially overnight - the programming framework that runs civilization.
Definitely been getting Idiocracy vibes the last decade or so, holy shit is it tangible right now.
This feels similar to the shift from writing everything from scratch to the use of massive libraries and frameworks. We were shifting from a paradigm with a single developer likely knew every line of code to three years ago when developers wrote the glue that holds frameworks together, but relied on the framework devs to support and test a large amount of the code base.
The difference here is that the devs are still responsible for AI’s code instead of the support of another company. What we’re going to see are more tools to certify the safety of AI developed software. Not only will software be tested against requirements, but also a stringent default set of tests to certify it against security risks.
The next 3 years are going to be a paradigm shift in software development. Things will be a bit messy in the interim. On the other side will be a new type of developer working with very different tools.
Not only will software be tested against requirements, but also a stringent default set of tests to certify it against security risks.
And diesel-gate tells us exactly how that’s going to turn out.
I’ll comment here, rather than the other discussion we were having.
To answer your question here, you don’t. Here’s why:
Do you farm everything you eat? Do you weave the textiles that make your clothes? Do you smith the shoes that go on your horse? Do you form the bricks that made your house yourself? Do you harvest fire wood to cook your food? Do you tan leather to make belts and shoes? Do you
No. And why? Because technological improvements made the artisinal version of that work unnecessary to be largely done by artisans. Not that we don’t still have artisans doing these things. And not that we have every single article of clothing on Earth made by machines.
Coders are in the same plight that cobblers were in 100+ years ago. And I’ve been to a cobbler in the last year! But only because I really love this one pair of great boots that can be resoled. I don’t go there for running shoes and flip flops.
The part that feels like betrayal to coders is that the LLMs came from their own kind.
I don’t farm what I eat, but the people that do have to have that knowledge of how to do that.
That’s the difference here. LLMs are creating “developers” that need to have the knowledge to be able to audit the code produced, but they won’t have that knowledge. That is the problem. Not the “paradigm shift”.
Not really.
LLMs do great at what’s expected cheap and fast. So for needs that are already well-worn territory, yes LLMs will do great. Do you need basic HTML, CSS and JS? Great, that’s a perfect use case. Easily handed by any LLM
Are you coding a bespoke system that is unique in an obscure language? LLMs will suck at that.
There’s a lot in between. It’s more about the bulk of people doing basic coding work where an LLM can get it 60% right that you’re taking about.
Ok, so here’s another example. You likely don’t bake all your own bread. People used to either bake their own, or get it from their local artisan. You can, and many people do, still bake their own bread. I bake maybe 20% of my own bread. Many people also buy from artisan bakeries with $8 loaves of sourdough. The skill was never lost, but the bulk of basic unfancy needs are easily handled by generic, basic bread.
And a lot of people are fine with buying garbage white bread made by a machine and never touched by human hands. You know what’s actually a lot of work to make? Plain-ass white bread, perfectly sliced. It’s a great time savings if that’s what you need.
I brew beer, too, and you know what’s a total pain to make by hand? Basic-ass light beer. I make stouts, and would never waste the time on a basic, $2/bottle light beer. But a batch of my own stout, vs. paying $8 a bottle for fancy stouts? Yeah, I’ll make it myself. Even if I make it with help from some malt extract, it’s still better and cheaper than artisinal beers, and it’s what I want for myself.
The vast, vast majority of people on Earth think ramen is nothing more than noodle packets with shit broth. I love making a 2-day tonkatsu broth. Really develop the flavors, simmer for 12-18 hours. Cook the egg just right… Love it because it’s worth it. But if I’m starving and need something fast… Maybe I would eat a cheap ramen packet.
With food, there are hundreds of similar examples. No one has lost the skills, but that also doesn’t mean that everything needs to be done artisinally all the time. We do this all the time in life. Just because this is the iteration that happened when you were an adult doesn’t make it unique or unusual. We’ve done this as a species often. Once a generation, or so.
A lot fewer people can make money being artisanal bakers these days, though. Used to be every village with more than a few families had a baker, all those bakers lost their bakeries and had to get different jobs working for someone else.
The same will happen with programming. There will be some people who still perform unique skilled work that can’t be slopped out, and then everyone else will need to get a different job. Every time this happens it causes economic crisis until the economy adjusts.
Did you just read my example? I literally just used this same example somewhere in the comments.
No one lost the skill. People interested to learn the skill can learn it. It’s just not a widely known skill because units you rant to know the skill, there’s no reason to know that into…
I just read the comment I replied to?
No one lost the skill […] It’s just not a widely known skill
It was a widely known skill, and now it isn’t. That means, quite obviously, skill has been lost.
Use it or lose it. If you don’t bake for 20 years and then get back into baking, you’ll find you have to relearn some stuff.
Skills get lost all the time when they become less critical. If you don’t need to know how to bake to have bread, why learn to bake?
Ah, OK - I think you’re conflating a few terms here, which is causing the confusion.
First, a skill that must be practiced and kept up is called a perishable skill. CPR is a perishable skill, you have to get re-certified to prove you’ve practiced. Speaking a language is also one. I used to speak Spanish pretty well and French OK, and over the last 15 years of not using either of those and learning other languages instead, my Spanish sucks and my French is limited to like 50 words. “Skill Decay” is the term for when someone used to know how to do a perishable skill well, and then stopped practicing and would need to practice more before doing it again.
Use it or lose it. If you don’t bake for 20 years and then get back into baking, you’ll find you have to relearn some stuff.
Right - that’s a perishable skill. If you experience skill decay and forget how to bake well, it doesn’t meant that the whole world has also lost the ability to bake bread.
Second, “Lost Skill” can mean a few things - I’m talking about it in terms of a societal lost skill. I don’t think you are, so it’s sort of a nebulous conversation pinned to a single term.
I’m talking about lost skills like hunting mammoths with an atlatl, or telephone switchboard operator. Skills attached to tasks where the task ceased to be part of our world. Mine horse trainer. High priest to the Temple of Zeus. Things like that.
What’s actually not a lost skill to our species? A lot of things where someone (or many someones) have kept the alive and recorded so we can still learn how to do it if we want. Papyrus scroll maker? You can learn that. How to make garum, the fish sauce that Ancient Rome used instead of salt on their food? You can learn that, too. How to hunt deer with an atlatl? We still have deer, and people still play around with atlatls, so it’s just hunting deer with a different weapon. Amazingly, there are people that hunt with spears for the hell of it. Weird, but they’re keeping a paleolithic skill alive in our species while others that are more modern have come and gone.
Back to bread. Why bake bread if you don’t need to know how? Personally, I learned how to brew beer, bake fancy bread, make pasta, make kimchi, make pickles, cure meat, and a few other things because I like those things and lived in places where the versions of those things I like weren’t available. The knowledge on how to do those skills was easily available, and so I learned in order to improve my quality of life. Many people do it for fun. Some people are lucky enough to do it for pay as skilled artisans that can charge $8 for a loaf of sourdough (getting up at 4:00am, notwithstanding the lucky part).
As for CS majors and people that write code, there are people out there that still know how to write COBOL and Fortran and early computer languages that used to be taught commonly. This is, so far, not a lost skill to our species, but it will be in 50 years as those people pass away, and no new programs are written in those languages. That’s like hunting with a spear. It might fade out and be lost, but it’s something people in the future can learn if they want. Modern coding is used everywhere all the time, so that skill in general is not at risk of being lost, even with LLMs.
When I was in high school, I wrote HTML by hand because my friends and I were dorks and thought using Word to get basic WYSIWYG formatting was “cheating.” We made fun of each other as sissies and bitches for using Word to format our Geocities pages - can you believe it? Then after maybe 6 months, we all stopped being idiots and realized that while knowing the basics were important to make little changes, letting the machine help you on the bulk of the work was just a huge savings of time and frustration. Once all our geocities pages were up and running, no one inspected the code anyway, so it ultimately didn’t matter other than our teenage egos.
Personally, I think a lot of the friction here is similar to the fading of the US auto industry. Americans hold a lot of self-identity with their jobs. And, these are very good-paying, highly-skilled jobs that are narrow enough they don’t allow for much transfer of the skillset to things like HVAC repair or plumber. So to go from $180K a year to being made redundant and learning that you’re now unskilled for most other work available is a double shock, and demand learning skills for another unrelated job entirely from scratch.
It sucks, but it happens a lot on this planet.
I feel like I see one of these posts every three months or so, and it’s just beginning to feel like meaningless doomposting, like the boy who cried LLM.
I don’t care how carefully the code gets reviewed, you don’t know it as well as if you had actually written it, and that will make all the difference three months from now when someone discovers a bug in your code (there’s always a bug in the code), or when the customer comes out of left field with a feature request that you didn’t forsee and isn’t easy to just add in (the customer always wants what you haven’t accounted for).
Don’t give up hope. Don’t let the clankers win. Continue to hone your craft. We will need senior devs more than ever after the AI bubble pops.
This is how I feel too. I’m in several work chats where everyone seems to be pretending that Claude doesn’t have its arm shoved up their ass, wielding them like a puppet.
Several of the projects I am working with are just AI generated. In fact, if I need to suggest any fixes, they will be denied unless it’s AI generated.
I’m tired. I’m depressed. I used to take so much pride in solving problems, but all of that is gone now.
I’m going to try to ride it out and see where it takes me, but the work is mind numbingly boring and the trades look better and better every day.
if it helps, they are completely unusable for specialised development.
Agreed. I’ve had a plethora of issues with very little improvements as time progresses.
What kind?
embedded systems, signal analysis, weird edgecasey lisp stuff…
I work in embedded systems and until recently I would have agreed with you… But it’s getting to the point where our agents have access to debuggers/tracers, system I/O logging and basically every tool we would ordinarily use and it’s vastly improving the quality of their work.
I always attributed their better performance in web and general software development to more availability of open source examples that they are trained on, but I think the ability to ‘self check’ was a huge limiting factor. Unit and other pure software testing abstracted away too much of the real system. With real feedback and access to the data sheets it’s doing a pretty good job.
Only four months back I told anyone who asked that LLMs would never replace developers, because back then they really couldn’t. The modern models in conjunction with their harnesses are a very different thing.
Similar time frame for me. I just spent a few hours documenting how a tool we already have should work (one that’s a freaking overcomplicated mess and impossible to manage) and feed those requirements to Argon. It made it 90% accurate in one shot. 5 % difference was an issue with my specifications and 3% a latent bug on the existing framework nobody had noticed.
Freaking scary
That’s scary. My most recent experience was a year ago when a coworker vibecoded a bunch of tests that somehow only tested the emulated hardware and forgot to test production code.
Try it again. In the past few months they are good enough for general enterprise dev that what OP said is the reality. We just watch our robots shit out a working solution to a ticket.
doesn’t really help you when any proprietary code crossing the network boundary is a security breach.
This is only a short term problem. More and more models are released that are „good enough“ that can be run on local hardware. Prices for components at the moment are a problem, but that’s a money problem not with the technology itself.
What does “good enough” mean here though? At the moment I highly doubt that we’ll get anywhere near the performance of frontier models on hardware accessible to the average user, due to the limitations of physics itself. Then on the model side of things I think there’s probably a limit on how small we can go before accuracy becomes a major issue, and I don’t think that limit would be something that consumer hardware would be able to run (and by consumer hardware I mean your average sub $1000 gaming PC, not bleeding edge M6 Macs)
I dunno man claude one-shotted a firmware build for an ESP32-based project I was working on last weekend. I had to massage it properly and I wouldn’t have been able to supervise it properly without my EE degree but it did produce valid and correct firmware.
Can you explain what you mean by “I had to massage it properly” when you one-shotted it? Did you preemptively account for things or something?
I mean, yeah. Asking “do xyz make no mistakes” won’t get you what you want. Come up with a plan, decide on what constraints matter, pick libraries you want to use. Decide what doesn’t matter. Generally around now I feed it my plan and ask it “do you think I’ve missed anything?” and read what it suggests and see if I agree with any of it.
From there, plan out what incremental progress looks like. Maybe start with the UART/I2C/SPI library and just reading firmware version off of each child component. Then ask it to go through the plan, increment by increment. Review everything as it does it, watch the console and redirect it if it’s making poor choices or just things you know will box yourself in in the future.
On my best day in university my project this weekend would have been weeks of work. It took me a day, not including soldering and packaging after the electronics were functional. It’s just undeniably faster at this point. Even if the only thing I used it for was to open the serial port, watch the logs, and fix any bugs I introduced, it would be faster. But you can feed it whole datasheets and it’ll construct a client library for your device essentially perfectly on the first try.
I have serious ethical qualms about how LLMs were created, how they’re going to be used, the inequality they create, the finances of the industry behind them, the environmental impact… but from a purely technical perspective they can be very powerful tools. But you need to already have domain knowledge to be able to massage it properly - “everyone is engineer” still isn’t reality. Someone with experience and good intuition though…
I work in driver developer and once I gave them access to the software verification platforms we use and the lab machines containing the test silicon, the quality improved drastically.
Most of my job is now interacting with people because those things can’t meaningfully turn business requirements into architecture yet.
it’s interesting how years of security hardening and multi-tier firewall infrastructure to prevent code leaks is all going out the window, isn’t it?
I’m not sure exactly what you mean by code leaks here. Like if the AI agent decided to take the proprietary code and post it on social media for some reason?
The same guard rails the humans had are in place for the AI agents too (not that they were super effective for humans to begin with).
anything the model ingests leaves the premises. that’s a leak.
I work for a very large company that owns their AI, so that’s not a concern in my case.
But even if I wasn’t, I would assume the company would have negotiated a contract with whoever provides the AI that meets their requirements. That’s above my pay grade.
I would never use a personal AI for company code. Even if public AI offerings aren’t blocked on corporate devices, that sounds like a great way to get in trouble for mishandling company IP.
You think they’re using in-house models to do their coding? 😂
signal analysis
I mean. You can claim that. And maybe its not via LLM, but if you are using LLM to use other models, it absolutely can.
i have no idea what you are trying to say.
I think “No AI” gets very myopically focused on "LLM"s but there are many other kinds of machine learning, and something agents are phenomenal at, is implementing basic code.
So sure, an LLM might suck at denovo prediction of sequence data. But an LLM can instead pull in a model that is quite good at sequence data.
Being against something doesnt mean we can be in denial of it.
i’m not opposed to machine learning in general, i don’t think anyone here is. i am myopically opposed to large language models. no other form of machine learning is causing the damage they currently are.
i’ve used machine learning techniques in my code for many years. llms are awful at seeing and applying those techniques.
If it helps², my colleagues are all too dumb to use it properly. If you yourself don’t understand what you’re trying to implement, it doesn’t matter how good your tools are.
I have the feeling it was never that necessary to review every piece of code than it is right now, because it is all too easy to write confidently without having a clue.
So far!
If you can direct them to good sources of information they can be usable. It’d be good if they were trained on researching rather that do X. As it stands, I’d rather lead horses to water instead.
As a (not so anymore) newly graduated computer engineer, this made me cry. Whenever I see threads like this the consolation thrown around is that someone with experience will always need to check the code and seniors didn’t used to write much code anyways. And juniors are all LLM jockeys anyways so who cares.
Well fuck me I guess, the last train to break into the industry departed 3 years ago when I thought I was learning my profession. A couple months ago I felt principled, I wanted a job where I wouldn’t be forced to vibe code. Now I’ll take anything.
At this rate even if I do manage to get hired, and even if by some muracle I’m allowed to write code there; the experience I gain from it wont be valued the same as someone who started working before LLMs became widespread. I’m a part of the slop generation now.
Makes me wonder what answer pr people want when they ask if and how I use ai. Do they want me to use them or not? If I use them, I accept that I have no place among the senior developers of tomorrow, if I don’t I’m missing on the one skill that companies looking to hire me expect me to have for the sake of something likely to be obsolete.
Maybe I shouldn’t have made being a developer a part of my identity while still in university. Is it time to let it go? Accept that I’ve spent years on something I wanted and believed would lead to a good life for it to just not pay off? That the safest most sensible fucking life choice imaginable just left me unemployable?
Maybe it’s for the best, as they say “Don’t turn your hobbies into a job”. But it breaks my heart to let go. Where would I even pivot here?
Sorry if this was rambling. This is a cry for help
This entire thread is pure doomerism. Don’t let it get to you. I’ve been in this field for just long enough to recognize this house of cards for what it is. I’ve seen the garbage LLMs spit out and anyone who thinks developing like that is sustainable either had their soul beaten into submission, or they’re full of shit.
I don’t wanna repeat myself, so I’ll link a previous comment: https://lemmy.blahaj.zone/comment/22227493
Stay strong friend.
For real, like all LLM output it’s good at looking convincing. It produces so much code though, and it’s not prioritizing efficiency… why would it, more tokens means more revenue.
Tokens are heavily subsidized right now, going to be a reckoning when they start jacking up the prices.
I’m not an expert, but maybe you can say you use LLMs for some projects and develop others by hand for the best of both worlds?
Do they want me to use them or not? If I use them, I accept that I have no place among the senior developers of tomorrow
Untrue IMO. A senior developer knows when to use which tool. As a future senior developer you’ll need to know when to use AI and when not to.
I see people who have been writing code and architecting systems for 20 years use AI to speed up parts of their workflow. More time left over for architectural work when 2 hours of debugging a stupid bug can be a prompt fired away while you go think about more big picture things.
If I had to interview for a job right now and was asked about AI usage, I would say that I’m definitely going to lean towards more manual work in the beginning to understand the projects and the tech stack better, and once I’m more comfortable with my knowledge, I’ll use more AI again, as I can verify its output. In the meantime, it’s still a fantastic search tool too.
If it makes you feel better they were selling you a bill of goods from the jump. Programming hasn’t been a stable or good career for ten years and the writing was on the wall twenty years ago. The fix was in when “learn to code” became commonplace.
The prevalence of the h1b program would have made your course of study a dead end in a few years even without llms.
Get an electricians certification and be a lineman. There will always be construction and storms. We’re building new houses so shitty that it’s guaranteed and there isn’t a weather control gun yet (even if there was, no one would use it to protect the people who live in tornado alleys or floodplains). You can’t get into that job in middle age so start young and avoid the temptation to buy a duramax, it’s how 20 year olds making 150k get fucked over. When everyone gives you shit for not having a truck, buy that old 6.0 and have a good mechanic fix it up.
If you really need the validation of typing instructions to a computer, go into embedded systems. Everyone will make fun of you because they’re twice your age but it’s a job. Those controllers aren’t going away because it’s a multi ten to hundred million dollar contract to replace them and it’s a few tens of thousands of couple hundred thousand to update them.
Stay in white collar work if you ever plan to get into it again. There’s a one way check valve on that so the moment you drive a truck you’re never getting hired for any entry level office position again no matter your qualifications. You can move back into office work at the management level though. The move is to make management on a work crew and then move back into an office.
Consider joining the brain drain. There’s lots of foreign companies happy to suck up discounted domestic talent.
Your work isn’t your identity no matter the job.
Being flexable in the coming years will mean having a savings you can fall back on and pumping it back up after you do. Learn how to set a budget and how to stick to it.
If people have to find a new career after they paid all that money for a degree, colleges should need to fully refund CS degrees.
LLM companies should also be taxed on tokens like labor is taxed. They also should have to retrain their models without the stolen, illegal data sources they used to train these models.
These companies just committed the biggest theft in human history and they’re not going to suffer consequences?
They need to suffer consequences.
Neither here nor there in the context of a reply to a self professed cry for help comment.
Of course neither you nor I nor the combination of every person either one of us has ever laid eyes on possesses the ability to perform the actions necessary to achieve what you suggest.
Even if what you said were possible it’s neither forward thinking or just, excepting a regressive, retributive conception of justice.
There was a report from MIT about AI and the friction of learning. And how there is no substitute for the friction. I think the friction is less with LLMs compared to figuring something out yourself. But that just means their needs to be another way of learning or friction for the job. People use to do a lot of calculations in their head, it was a valuable skill, for me, I just learnt algorithms as more abstract concepts that can translate into a programming language because that was the valuable skill. Learning what’s next will look different, companies may not understand, but the ones that do will last.
deleted by creator
Additional thoughts: deepseek is running a 3x profit on their gpu/hours so there is no salvation in a bubble pop (we’re not getting that, the whole us economy is built on ai now so when things go south it’s gonna be commercial and residential real estate and energy conditions that counteract the effects of a pop and lead to much more controlled deflation instead) because it’s already worth it and economically sustainable to develop at deepseek capacity and prices, which if you have not tried it is a fucking hoot.
So anyone telling you to hold on and wait for the pop doesn’t know what they’re talking about.
It was a meme a few years ago but look into public sector work and old obsolete languages. If you have t got a pass through llc and tax id you can use to pick up contracts then get on that. It’s like ~$200 a year and makes you safer and able to mass apply for contracts.
Do not keep trying to do everything by hand. Learn to use some kind of harness in your dev environment. You probably got trained on vscode with its tab completion. That’s ai. You’re already using it.
The job hasn’t died. How you do it has. Your job was never to write code. Your job was to understand problems and transform data.
The senior dev’s job was never to simply write code. But the junior dev was made to handle the grunt, code-monkey work while picking up senior dev skills and experience.
Now you need to have senior dev skills to just get work.
Once upon a time the junior dev’s job was to clean the mainframe and type up the punch cards.
It used to be that airliners had a captain, co-pilot, and engineer to manage the engines and other systems. With increasing automation and system reliability the engineer position was made redundant. No profession is safe from stagnation.
Can do you one better.
Automation Dependency is a legitimate worry. Aviation has also been safer post-1997 than it was pre-1997.
Once upon a time the junior dev’s job was to clean the mainframe and type up the punch cards.
Source? Because I worked for a utility that used to use punchcards. I actually had a card reader and paper tape drive behind me in the server room where I worked. The old timers would come in and reminisce because it beat working, and none of the guys I talked to let “junior” devs touch the punch cards.
Imagine if they let the new guy waste a bunch of cards to put the machine in an unworkable state, or fuck up the punches from someone else’s code, or – and this was a very real problem – drop the entire box on the way to the computer. The cards were srs business and the concept of “junior” didn’t really exist because computing was a brand new field.
No profession is safe from stagnation.
Yeah, yeah. No profession is safe from short-sighted leadership taking a short-term gain instead of growing their talent either. That’s not something to celebrate.
Different shops had different practices, who knew!
It was seen as training on how to use the machine and access to what the more experienced coder was doing. It was of course also back when on-the-job training still existed instead of just throwing the new guy in a corner and expecting them to pick it up themselves.
Once again: source?
deleted by creator
Same as yours.
Second point still stands.
Yeah your job is solving problems with computers, by knowing how to send instructions to a computer to get it to reliably produce desired results.
making programming about writing code is like making photography about film development and darkroom printing. you spent years perfecting those skills and learning about things like effects of different chemicals and dodging and burning amd then the digital camera comes along. and you’d be like “anyone can make a picture now, they edit them on a computer, photographer as a profession is dead”. it’s pathetic.
