alt Hacker NewsVibe coding creates fatigue?
Comments
There probably needs to be some settled discussion on what constitutes "vibe coding." I interpret this term as "I input text into $AI_MODEL, I look at the app to see my change was implemented. I iterate via text prompts alone, rarely or never looking at the code generated."
vs. what this author is doing, which seems more like agent assisted coding than "vibe" coding.
With regard to the subject matter, it of course makes sense that managing more features than you used to be able to manage without $AI_MODEL would result in some mental fatigue. I also believe this gets worse the older you get. I've seen this within my own career, just from times of being understaffed and overworked, AI or not.
I've felt this too as a person with ADHD, specifically difficulty processing information. Caveat: I don't vibe code much, partially because of the mental fatigue symptoms.
I've found that if an LLM writes too much code, even if I specified what it should be doing, I still have to do a lot of validation myself that would have been done while writing the code by hand. This turns the process from "generative" (haha) to "processing", which I struggle a lot more with.
Unfortunately, the reason I have to do so much processing on vibe code or large generated chunks of code is simply because it doesn't work. There is almost always an issue that is either immediately obvious, like the code not working, or becomes obvious later, like poorly structured code that the LLM then jams into future code generation, creating a house of cards that easily falls apart.
Many people will tell me that I'm not using the right model or tools or whatever but it's clear to me that the problem is that AI doesn't have any vision of where your code will need to organically head towards. It's great for one shots and rewrites, but it always always always chokes on larger/complicated projects, ESPECIALLY ones that are not written in common languages (like JavaScript) or common packages/patterns eventually, and then I have to go spelunking to find why things aren't working or why it can't generate code to do something I know is possible. It's almost always because the input for new code is my ask AND the poorly structured code, so the LLM will rarely clean up it's own crap as it goes. If anything, it keeps writing shoddy wrapper around shoddy wrappers.
Anyways, still helpful for writing boilerplate and segments of code, but I like to know what is happening and have control over how my code is structured. I can't trust the LLMs right now.
I feel this.
I take breaks.
But I also get drawn to overworking ( as I'm doing right now ), which I justify because "I'm just keeping an eye on the agent".
It's hard work.
It's hard to explain what's hard about it.
Watching as a machine does in an hour what would take me a week.
But also watching to stop the machine spin around doing nothing for ages because it's got itself in a mess.
Watching for when it gets lazy, and starts writing injectable SQL.
Watching for when it gets lazy, and tries to pull in packages it had no right to.
We've built a motor that can generate 1,000 horse power.
But one man could steer a horse.
The motor right now doesn't have the appropriate steering apparatus.
I feel like I'm chasing it around trying to keep it pointed forward.
It's still astronomically productive.
To abandon it would be a waste.
But it's so tiring.
> One reason developers are developers is the dopamine loop > You write code, it doesn’t work, you fix it, it works, great! Dopamine rush. Several dozens or a hundred times a day.
This statement resonates with me. Vibe coding gets the job done quickly, but without the same joy. I used to think that it was the finished product that I liked to create, but maybe it's the creative process of building. It's like LEGO kits, the fun is putting them together, not looking at the finished model.
On the flip side, coding sessions where I bang my head against the wall trying to figure out some black box were never enjoyable. Nor was writing POCOs, boilerplate, etc.
This morning I attended and paid attention to three separate meetings and at one point had three coding agents running in parallel solving some quite complex problems for me.
It's now 11:47am and I am mentally exhausted. I feel like my dog after she spends an hour at her sniff-training class (it wipes her out for the rest of the day.)
I've felt like that on days without the meetings too. Keeping up with AI tools requires a great deal of mental effort.
Going fast is awesome. If you have a bit of caffeine in the morning, turn on some tunes, and get into your workflow it's awesome. I get so sucked into my coding projects that when I'm done I'm disoriented. Nothing quite like being in the zone.
I see people with no coding experience now generating PRs to a couple of repos I manage.
They ask a business question to the AI and it generates a bunch of code.
But honestly, coding isn't the part that slowed me down. Mapping the business requirements to code that doesn't fail is the hard part.
And the generated PRs are just answers to the narrow business questions. Now I need to spend time in walking it all back, and try to figure out what the actual business question is, and the overall impact. From experience I get very little answer to those questions.
And this is where Software Engineering experience becomes important. It's asking the right questions. Not just writing code.
Next to that I'm seeing developers drinking the cool-aid and submitting PRs where a whole bunch of changes are made, but they don't know why. Well, those changes DO have impact. Keeping it because the AI suggested it isn't the right answer. Keeping it because you agree with the AI's reasoning isn't the right answer either.
A guy at work did a demo of an agent work flow for some higher ups (we have chatbots but haven't adopted agents yet). He raved about how after writing a several hundred line spec, being extremely specific about the technology to use, and figuring out where to put all the guardrails, he was able to get Claude to generate weeks worth of code. When all was said and done it was like 20k lines of code between implementation, tests, and helper tools. Along the way he acknowledged you have to keep a close eye on it, or it will generate functions that pass tests but don't actually do their jobs, tests that pass but don't test anything, and a bunch of other issues.
To people with little to no practical software experience, I can see why that seems incredible. Think of the savings! But to anyone who's worked in a legacy code base, even well written ones, should know the pain. This is worse. That legacy code base was at least written with intention, and is hopefully battle tested to some degree by the time you look at it. This is 20k lines of code written by an intern that you are now responsible for going through line by line, which is going to take at least as long as it would have to write yourself.
There are obvious wins from AI, and agents, but this type of development is a bad idea. Iteration loops need to be kept much smaller, and you should still be testing as you go like you would when writing everything yourself. Otherwise it's going to turn into an absolute nightmare fast.
Claude Code latency is at the unfortunate balance where the wait is long enough for me to go on twitter, but not long enough to do anything really valuable. Would be more productive if it took minutes or under 5-10 seconds.
This is why I document my ideas and then go for a walk. It also helps me stay within the quota limits.
I find vibe coding similar to visiting a country where I don't know the local language very well.
Usually that requires saying something, seeing if the other person understands what I'm saying, and occasionally repeating myself in a different way.
It can be real tiring when I'm with friends who only speak the other language so we're both using translator tools and basically repeating that loop up to 2-3 hours.
I've found the same situation with vibe coding. Especially when the model misunderstands what I want or starts going off on a tangent. sometimes it's easier to edit the original query or an earlier step in the flow and re-write it for a better result.
Same for me. I feel fatigue and I have also been reflecting about that as it reminded me the same type of mind exhaust I felt when I was a manager.
Choosing, analysing, verifying, talking to others rather than thinking in a sequential and organised way.
One of the reasons I moved away of that position was this constant fatigue by the end of the day. I was happy on seeing things moving forward but it felt that it took more energy to do less
Downtime for the conscious brain is paramount in life, as that allows the sub-conscious to absorb and act on new information. I have no science to point to, but I believe wholeheartedly that the conscious and sub-conscious minds cannot access the same neurons at the same time, it's like single-ported RAM. More than one thing in my life has been improved by taking a conscious break and letting the subconscious churn
While the productivity can be exciting. I personally observe the sa.e fatigue as described. I chock it up to the speed or frequency of context switching as a result of faster implementation.
No; stupid terms like "vibe coding" create fatigue.
There’s a lot of points I agree with but I think what’s important is fully conceptualizing the mental model of your project. Then, context switching doesn’t even induce much mental fatigue.
true in my case. when i get into the zone while coding, i can go on and on. while llm can help, there's a cognitive mismatch between coding directions to llm and reading the code when it comes to continuing coding. brain and the generated code aren't aligned. i prefer at the moment, after coding a feature, to see if it can be improved using llm. and it's of great help to write tests.
In my experience, it depends on the task. As a hobby, I develop personal projects (mostly web apps) where I'm not an expert in the relevant technology. In this case, LLM-assisted coding is empowering - and I don't think it's Dunning-Kruger, as I can understand the generated code, and it "feels" good enough given my 20+ years' experience in software engineering.
However, when it comes to my professional work on a mature, advanced project, I find it much easier to write the code myself than to provide a very precise specification without which the LLM wouldn't generate code of a sufficiently high quality.
Doctorow's "Reverse Centaur"
This is what loom workers experienced after the introduction of the power loom, what factory workers experienced under Taylorism, what Amazon workers experience today in many cases. Just working at a pace that is unsustainable. This is why unions exist.
I play stupid web games (generals.io I'm looking at you) while Claude does its thing. Takes the edge off the pace a bit.
This fine for WFH/remote work. It didn't have great optics when I went back to in-office for a bit.
I find the increased speed to be an opportunity to slow down and focus on certain things, particularly the architectural changes of the feature I'm looking to introduce.
If you're able to blaze through feature tickets using GenAI on existing projects of any major complexity, there's almost certainly something which would produce better code that you're skipping.
I have plenty of info and agents for Claude Code to take into account when I use it to make features in our projects, but what it can't know is the cadence of what our business partners expect, the unknown unknowns, conversations that humans have about the projects, and the way end-users feel about the project. My job is to direct it with those factors in mind, and the work to account for those factors takes time.
If you're automating code generation but not automating verification, there will be a mismatch.
Maybe the fatigue comes from that mismatch?
The classical vibe coder style is to just ignore verification. That's not a good approach as well.
I think this space has not matured yet. We have old tools (test, lint) and some unreliable tools (agent assisted reviews), but nothing to match the speed of generation yet.
I do it by creating ad-hoc deterministic verifiers. Sometimes they'll last just a couple of PRs. It's cheap to do them now. But also, there must be a better way.
anecdata
I don't want to be that contrarian guy, but I find it energizing to go faster. For example, being able to blast through a list of niggling defects that need to be fixed is no longer a stultifying drag.
I recently used a coding agent on a project where I was using an unfamiliar language, framework, API, and protocol. It was a non-trivial project, and I had to be paying attention to what the agent was doing because it definitely would go off into the weeds fairly often. But not having to spend hours here and there getting up to speed on some mundane but unfamiliar aspect of the implementation really made everything about the experience better.
I even explored some aspects of LLM performance: I could tell that new and fast changing APIs easily flummox a coding agent, confirming the strong relationship of up-to-date and accurate training material to LLM performance. I've also seen this aspect of agent assisted coding improve and vary across AIs.