I was using tabnine in 2020, I used copilot autocomplete + ChatGPT, then Claude Code. All this to say, I willingly adopted each wave.

Having said that, for all our startups “interesting” code I use no AI, not even tab complete. For a few reasons:

- having derived the code and built a mental model is something like 90% of the work, the code artefact being 10%. (You can probably verify this yourself, if you’ve implemented something once and then someone deleted it all, you could rewrite it MUCH faster the second time).

- once I have this model / vision in my mind, I’m streaming it from my mind into reality via code and having code suggestions pop up breaks that flow state for me.

- using Claude code even on throwaway scripts can mask problems. Why is doing this thing so difficult that it needs an AI to write the code? And usually when I’m writing those scripts I have a thought or insight about the system. in other words it’s nice to have that low stakes time.

- models are still frustratingly bad at spatial reasoning and I don’t see this improving anytime soon, which is a problem for a lot of graphics code.

- the code models produce still isn’t what I’d consider great code; subjective of course but it’s our product and we think having great code in our core technology is well worth it in the long term!

I do still use Claude code on various commoditised pieces (account UI, etc) and I use Claude as a great search engine, though its explanations of complex topics still kind of sucks and a human blog is almost always better.

We are a small company of 15 engs.

In my startup, I am currently in consideration for doing this exact thing. The interesting part is it's not really the AI that's causing the issue. It's the ease of my developer to make a mess, either with messy code or feature bloat.

After figuring out the exact product we need with some very quick development cycles with AI, we now have some problems that I can't keep under control because of AI. So, currently under consideration to rewrite the core functionality without AI so we can keep it simple, understandable and slower to change.

The interesting part is that AI could do this, but managing developers with AI has become very difficult to get them to slow down and build stable simple things.

EDIT: I found it interesting how negative the comments were and I thought it would be easier to add context to the root.

My developers wanted to use AI. I was a skeptic for several reasons. But I agreed that it was worth committing to and seeing what came of it. And all in all, the rapid prototyping is excellent.

However, as parts of the software are past the point of rapid prototyping, I talked with my team and everyone agreed that the core piece deserved a re-factor/re-write because of the mess it was. Classic tech debt of deprecated code hanging around and prototype features with patchwork additions that now are a poor way of executing the intended function.

So, we agreed to do a significant re-factor and did. However, quickly it got incredibly messy again. Partially due to AI using git to look at prior commits and add things back.

And so, it was essentially an unsuccessful refactor and we find ourselves with the same issue. They also agree that using AI is the problem here because it's very easy to get new things. And the temptation to solve a problem quickly is high (I am stressing the opposite, to go slow), but it can be hard to say no to a co-worker who isn't dealing with the technical issues.

So, both myself and my developers agree that AI is the issue in creating a messy that has resulted in unstable code. The reason I say that I am considering this choice is because while they acknowledge the issues it's causing, they aren't excited about returning to coding without them. Which is also why I'm considering the right move here, since the goal is both a happy team and well functioning software.

Yeah I see the same thing. Also shows up when people produce an “analysis” largely driven by AI and they haven’t really thought it through. They/we are outsourcing their thinking.

That's what affecting me the most. Pure Claude hallucinated analysis that the poster has not really read; worse part is that I can ask Claude for the same analysis, so there's no benefit, only long questions that waste my time.

> the goal is both a happy team and well functioning software.

software first. team second.

team happiness is important as a miserable team will hamper productivity and quality. but the team’s happiness is secondary to their overarching reason d’être — the software.

if there were no need for the software, the team wouldn’t need to exist. so that goal should take precedence.

YMMV

I think I definitely see the problem as a solo dev. I can just wing it with refactors and new tests, and all manners of sloshing about. But if I tried to do this with just one other person, they'd be pulling their hair out trying to understand why the last time they had to understand the code has completely shifted.

There definitely going to be a lot of fortune 500 companies absolutely wrecked by the amount of changes to their code base that no one will be "responsible" for because none of the workers will really know if it was their agent that changed something.

There could be some outliers who found the correct tools at the right time, but I don't think they'll suddenly be thrust into the money, because good code is definitely not the only thing that lets people succeed.

Check out AI-DLC or another structured development workflow for your team and stick to it. If you’ve got devs: not using any structured workflow, or using a variety (speckit, openspec, superpowers, their own) it’s going to be hard to control the quality of the development process itself and its outputs: you are just producing slop.

What many businesses are going through is the industrialization of software development.

Of course. You've described a company with no team.

What you've said also makes clear why nobody would have an incentive to play like a team player. You have this comprehension work you think is important and you've already devalued it by saying that humans would be a waste if AI could be used.

If I were a human working for you I would conclude that regardless of your exact words, your actions create the conditions whereby it is far safer to try to get the AI to bullshit its way through things than to risk trying to do them well as a human. You clearly articulate here that the bullshit is all you care for. If that is what will satisfy you most readily, why would anyone working for you aim higher anymore?

> The interesting part is that AI could do this, but managing developers with AI has become very difficult to get them to slow down and build stable simple things.

That is mutually incoherent. If AI can reliably do it, developers who reliably did it before AI would reliably did it again.

Sounds like AI cant reliably do it and your developers struggle to control its outputs.

> I am currently in consideration for doing this

Try “I am considering”.

Rereading, I can appreciate the active verb you want me to use, but I like "in consideration".

I am in a period of consideration, it's more passive which accurately captures the reality.

In a situation where you are an editor and focusing on the clarity and the brevity of the writing, I think this is a fantastic suggestion.

Hacker News has long ago decided that making comments on the quality of the writing in the comments is not welcome. Correcting spelling, correcting word usage, changing phrasing, and making other suggestions to improve the grammar of a post is not something that Hacker News readers want to encounter in the comments section.

It bothers me every day because there are gobs of cases where the writing is horrible, but I understand why the rule exists. It would make the comments section insufferable if everybody were just correcting each other's grammar the whole time.

Why are you being so friendly? Perhaps En isn't their native language. Although, taken out of context, that sentence could be interpreted as "I got myself in trouble for doing this". :)

This is correct. It's a giant red flag in GP that they're worried about their developers making a mess with AI. That's an issue of the developers' judgment not being any good, not a problem with AI.

If your developers aren't focusing on what's important to you, that's a leadership problem, not an AI problem.

I guess also useful to answer your question, is look at the companies who had a successful track record, of delivering reliable, consistent software over 5 to 10 years, went full vibe coding and got into a massive end crash.

You have evidence video here, that the AWS AI Division (04:23) went literally full vibe coding:

"I was laid off from my AI software engineering job at Amazon" - https://youtu.be/ZAom6f4-vh0?t=263

The whole division has been dismantled...

Ford did: https://www.forbes.com/sites/joetoscano1/2026/06/30/ford-hir...

So did Commonwealth Bank of Australia and IBM: https://www.cnbc.com/2026/07/01/employers-who-laid-off-worke...

And I'm sure many others who didn't publicize it because they have to keep this circular economy going.

The original question was about developers going back to manually coding without using AI. The link you posted is not about it. IBM replaced its Human Resources team with AI and hired them back. They didn’t remove AI usage in development and went back to manual coding.

Please read your sources completely before posting them only reading the title

The question was whether some companies forego LLMs in favour of human authoring

The original question was certainly specific to coding. It's right in the title.

I think it's funny that in the same thread I'm being accused of not reading the headline and reading only the headline

I was surprised to see IBM on that list

me too, till I read the article linked. As far as I know IBM actually encourages using AI, they have a really good youtube channel (ibmtechnology) that explains AI concepts and usages

Ford did not. They hired back veterans to better train the AI.

> COO Kumar Galhotra said Ford had been over-relying on automated quality systems without getting results, per Bloomberg. The returning engineers rebuilt the data pipelines feeding Ford's AI training, mentored junior staff, and reprogrammed the automated systems they had originally been brought in to replace.

Similarly with Commonwealth Bank and IBM, which are cited in your second link. None of these companies are saying they're not going to use AI and they're going to go back to manual labor. What they are saying is that they laid off workers prematurely.

The big lesson is highlighted well by IBM's head of HR, which is you can use AI, but you have to continue to invest in humans:

> “If we don’t continue to invest in entry-level hires, what happens in three-five years?,” IBM’s chief human resources officer, Nickle LaMoreaux, said at a Charter AI Summit in New York. “There’s no pipeline; the well simply dries up,” LaMoreaux added.

> circular economy

I get what you mean, but in general circular economy means something different, positive in my opinion. Here is what wikipedia has to say:

> a model of resource production and consumption that involves sharing, leasing, reusing, repairing, refurbishing, and recycling materials and products, to extend product life cycle for as long as possible

In the EU it is a hot topic for a lot of non-profit and social innovation work

I think the term GP was looking for was "circular financing", which has far fewer positive connotations.

Right. Almost every economy is circular in some fashion. Every producer is also a consumer. Sometimes the circle is quite big, but many of the dollars (or other currency) that you spend, come back to you in the money you earn.

Rehiring large numbers of people after AI-motivated layoffs is not the same as getting rid of LLM-generated code and going back to hand-written code as a whole.

No they didn’t. Did you read either of the things you posted?

We don’t have a no-AI policy but I’m trying to hand roll some code, because it is useful for interviews. It is very difficult to memorize the exact syntax unless I hand write code everyday.

What's your grade?

Senior. So I still get coding tests. I wouldn’t worry if I’m Staff and above.

My company hasn't switched at all... does that count?

I think there are many companies that have not hopped on the AI bandwagon yet. This is why AI companies still see a huge potential for continued growth.

Some are hesitant to chase after the latest fad (fearing it will soon fade); others want the industry to shake out many of the flaws first.

We saw the same thing with the Internet. Many companies (including some really big ones) were slow to embrace it until everyone else was on board.

What a dream! Don't you have investors that need hyping up all the time?

I don’t think that would ever happen. AI as a tool made the life so much easier removing initial inertia.

I agree that we still haven’t figured out what’s the best way to use AI tools, but over the time it would mature and people would come up with patterns, conventions, and design principles so the quality of what is produced will be maintained or increased while keeping the productivity gains.

Yeah I cant imagine my job without it anymore, in my hobby projects I dont really use it because its more about learning new stuff but for work I dont really care

The key part is “but for work I don’t really care”

No! the key part is "they dont't care why should I" ?

That's so depressing, and a real loss for your employer. Even if they neither know nor care that you are now sandbagging them, you still are. You aren't making real gains because you're tossing out value of equivalent or greater worth (your learning and engagement)

Most places I worked at cared very little for quality code, there was much more pressure on shipping fast even if it meant incurring technical debt, crap performance for end users, or developers leaving due to legacy code accumulation that nobody understood anymore.

I'm glad there is a tool that let's these companies have even shittier code shipped even faster. The faster they burn down the better.

This seems the right place in the discussion to mention The Gervais Principle, a legendary piece of business writing if you haven't heard of it. https://ribbonfarm.com/2009/10/07/the-gervais-principle-or-t...

You can only learn and engage so much in a day. In my head, it can be healthy to shift from tiring yourself out via work to tiring yourself out at home. You're not sandbagging them, you're setting boundaries on how much you're willing to do. If this is a problem for the employer, they can act accordingly.

But that's obviously exactly what the employer expects, I don't see the problem tbh ;)

I'm with you except that the economy will crash when it turns out everyone stopped creating real value. ...and in the mean time every real product I have to use is getting worse.

I think we're way past this point tbh, software development is just the latest area that's being "bullshittified" ;)

(also let's be honest, building yet another CRUD app - eg what 99% of software devs actually do, is not "creating value" - with or without AI)

Where they created value wasn't just the CRUD app but developing their skills, supporting communities that asked and answered questions to build public knowledgebases. They adapted to painfully repetitive tasks by developing tools that made expressing common ideas simple. In losing the pain of repetition, we lost the incentive towards abstraction. The forces in opposite directions cancelled out, creating growth for AI companies but overall stagnation due to the immense amount of value lost because before AI there was never a dollar value put on it.

It will happen once everything breaks and we have an entire generation of drooling idiots spamming prompts. AI is 100% trash and it will only get better at deceiving you into thinking it isnt.

Every single person I talk too that is enthusiastic about AI is so because they can't code for shit. Sure, then the AI seems better. But it never really is. Learn to code, that inertia you are talking about is the technical debt. It will drag you down into the abyss eventually

Can you understand that there is a lot of DIFFERENT people than YOU who can use AI tools? My team leader (he is also director of department) can use Claude/Codex (or other LLM/AI tools) and create faster, better things than doing it manually. You would not say that he is going down - he is even better with it.

You sound like an angry old man yelling at a television. I can code and have been doing it for 15 years before AI. If you don’t like AI that’s totally fine and don't be dogmatic, but then again the people who doesn’t change with time are left behind in time, maybe that’s why you’re so angry

One think I think is strange about the m programming world we live in now is the mythologization of pre llm code. As if people weren’t copying code from stack overflow, auto completing their way through APIs using intellisense, intellicode, tab nine, or checking in code there didn’t properly test or understand

There is a big difference between intellisense and LLMs. I've never used intellisense to input something I wasn't already looking for/about to type out. It just saved me remembering the exact wording and/or keystrokes.

My employer is encouraging AI adoption across the board, to the point of basically making tokenmaxing part of performance KPIs. I think they're pushing headlong into disaster, but they are not the kind of people to take advice, or admit mistakes.

I use AI coding tools both at work and in my hobby projects. It very clear to me they're nuclear-powered footguns, and we have a long, long way yet in developing practices, structures and workflows that will enable true benefits while minimizing the absolutely toxic baggage/fallout/side effects.

I look at the big corp push for AI adoption as another instance of boards of directors choosing the large investors / their own personal gain over the success of the corporations they govern (first instance being the "back to the office" push in attempt to shore up commercial real estate). I think the circular economy of the AI bubble has spread far and wide, and we're watching a lot of invested players trying to keep it going.

Not for code, but my wife is a doctor and has gone from using AI scribes back to manually typing out notes and has multiple friends who've done the same. Turns out the gains were eroded by having to check the work and turn verbose prose into an actual note.

> Turns out the gains were eroded by having to check the work and turn verbose prose into an actual note.

LLMs are great at this. I'd be very surprised if it wasn't possible to improve the actionable notes.

Most human-made doctor notes I've seen have not been comprehensive, anyway, and the doctor has had to spend a good chunk of a rushed appointment scribbling notes rather than interacting with the patient.

Exactly. It fascinates me when I'm emailed long AI transcripts of meetings with the disclaimer "generated by AI. Be sure to check for accuracy".

Like, did somebody seriously think through the meaning and implication of that disclaimer and still write it?

I'm running a startup and we deliberately don't use AI for development. We only use it for review.

We're working in a really deep area where having full understanding of our code is more important than speed.

Besides, in this area, clients getting one whiff of AI code would be an immediate deal-killer

> Besides, in this area, clients getting one whiff of AI code would be an immediate deal-killer

If you implement AI code review suggestions then you're using AI code, just the hard way.

Also I don't see why your clients would be ok with having AI review and influence your code. Seems hypocritical, or like you're trying to cheat them.

That's a fair challenge. AI review can mean many different things though. We treat it like a slightly better CppCheck, for context. TBH the way we use it doesn't really add much beyond regular static analysis tools.

Re "the hard way", I disagree. While it may seem like the hard way, in the medium and long term this is turning out to be the easy way. We have basically no technical debt, we can fix issues rapidly and (intangible I know) we can stand behind our codebase with absolute confidence.

Out of curiosity - what space are are in?

In what context would generating code faster not be a desired outcome?

Assuming of course everything else stays the same (quality, etc.)

I think that question hinges on the second part of your comment and should be answered as:

Does anyone know any company that went back to hand-written code because it decided drawbacks of ai generated code or some other concerns outweight the output benefits?

Lines-of-code has long been accepted as a terrible measure for code quality, and I believe it has been strongly correlated with poorer code quality.

No one is advocating for lines of code here.

It has a 1-to-1 correlation, though: the faster you can generate code (which is the one quality under discussion), the more lines of code you can generate in a given unit of time, and vice versa. So I would argue we are talking about lines-of-code, just indirectly.

I don't see how one can hold "speed of code generation" as a good metric while, simultaneously, holding "number of lines generated in a unit of time" as a bad metric. And more worrisome, I'm not sure how one would reward one without rewarding the other creating a perverse incentive. But I'm open to be proven wrong.

It's so weird how "assuming" is like watching super man hold a train back from a little girl on the railroad tracks.

Shitting out lots of code faster was already possible before (just hire more interns who do it for free heh) the question is at what quality and is more code in less time really the biggest problem in professional software development that had to be solved?

LLMs can be very useful for all sorts of things, but large scale code generation is IMHO the least interesting use case. And even when LLMs are successfully used for code generation, I feel like progress has been reset to the early 60s and people are now rediscovering all the failed software development approaches (eg "spec-driven" is pretty much the equivalent of "waterfall", I'm now waiting for UML diagrams to make a comeback as the next big thing with an "agentic engineering" label slapped on ;)

Classic interactive "vibecoding" with quick turnaround times (but much quicker than now please, don't make me wait and let me slip out of the flow) might actually turn out to be the most useful way to build software with LLMs (or let's better say "prototypes"). Because everything else currently looks like we're building up too much bureaucracy around the software development process again, and just in time when we finally got rid of that shit (now we have that absurd amount of .md files in pseudo-human-language as "skills", "rules", "context", "memory", ...) like it was common in the 70s when the work was split between "software architects" who only do the high level design, and "implementers" who only bake that design into code. This was obviously a stupid idea and I don't know why the "AI bros" seem to be so keen on repeating that mistake.

Every approach that builds a "human language specification" that's separate from source code written in a much more precise programming language is doomed to fail (eg the code is the spec!), and that's nothing new, we've known this for decades, but that brain virus of "upfront software architecture" always keeps creeping back into the minds of people at the first opportunity.

I think Spec-Driven Development is similar to waterfall but it is not like that. It is more agile oriented and for fast iterations. The idea is simple and can be vibecoded too - quick chatting, planning and so on.

The main things I don't like about LLM is that frontier models are paid (so no unlimited "tokens"), subscribtion based payments, local models are "not ready yet".

It's fascinating how people jump straight into their biases, making all sorts of inferences that were never mentioned.

Where did anyone advocate for "large scale code generation"? LLMs are a fantastic way to go from, "We thought this this feature" to "It's shipped and in people's hands". That could've been 100 lines or 1,000 lines but that's not the point.

You're thinking too small! It's all about building entire apps autonomously from 10 lines of a blurry "do what I mean" spec /s

I did witness ai output advantage dwindle as codebase grew in a team of experienced devs who switched to 99% ai generated changes.

The last time I've checked ai still had an edge adding new features, but the team collectively lost the project knowledge and any problem discovered that llm could not fix took significantly longer to correct - they were effectively working on a new to them codebase.

Do you know any company that went back from local AI to centralized AI?

the number of companies using local AI for development is too small to draw any conclusions from.

Best use cases are along the periphery: security checks, performance checks, test generation.

tests generated by ai, especially which generates each time codex are completely useless and needed only for beautiful "44/44 tests passed"

I think devs themselves won't want to work for a company that doesn't let you use Codex et al.

I dunno, I have friends and family at Meta, Roblox, $JOB, etc that would all prefer a no-AI environment. It's definitely not a majority of devs, but it's not an insignificant amount either.

I disagree. Personally, I would work in company which respects more code/infra quality than coding speed. When you vibe-deliver a lot of features per sprint after some time codebase becomes an unpredictable AI Slop.

> I would work in company which respects more code/infra quality than coding speed

Same. But that doesn’t have to be in contrast to using AI tools. There’s a lot of grunt in day to day coding as well.

The problem is not the tools

You absolutely can deny long term productivity gains.

Pre AI it took 100 engineers 5 years to get into a legacy code situation. Once you’re in a legacy code situation it’s very hard to add new features and your code is full of bugs. Fortunately most old companies with legacy code are making loads of money so they can pay the increased development costs to add features to their legacy code.

Now with AI five engineers can build a legacy codebase in six months.

Why five? What do the other four do?

Amateur numbers. I can build a legacy codebase in an afternoon

Neither can anyone deny that 'productivity gains' are worthless if what's being produced is worthless.

If you are producing worthless code, you are choosing that outcome either by choosing to work on worthless projects, or by failing to provide the direction and constraints that cause the model to create performant, maintainable code. You can’t just yolo everything, it’s still software development, but with LLMs it’s more of an engineering and management role.

I do. I deny the "gains". The "gains" will turn out to be a mirage in the medium to long term. The technical and cognitive debt incurred will be too extreme.

> Although code quality may sometimes suffer

That doesn't sound like a good thing in the long term.

> no one can deny the productivity gains AI brings to software development.

What are those "productivity gains"? Rapidly building hundreds of the wrong things that people do not want?

It means you cannot stop and the competition only moves far more quicker and its a forever race to the bottom.

Someone answered the question with some actual examples. Please refrain from the hyperbole

And those examples were not cases where companies went back to manual work after using AI. They were examples of companies regretting laying off their experts.

The point GP is making, which I understand seems dismissive, is that AI is a tool that people are not retreating from, because it offers the same sorts of gains as previous tool improvements we've seen in the field: from compilers to IDEs to sites like Stack Overflow. I think adding AI to that group of improvements makes a ton of sense, as it has the capacity to provide similar gains in productivity.

Exactly. It's quite scary that some people are reacting viciously to this type of question. There's real insecurity here...

Those things don't hallucinate. (Of course, stackoverflow can be wrong but at least its peer reviewed)

The best part of the downvoting of this is the answer to all of these is yes. Companies have banned all of these in the past when they were coming up. If you don’t learn from the past, …