Those companies are certainly writing more code. But It isn’t clear that they are increasing their economic productivity. It could even conceivably have the opposite effect by fueling a race to the bottom. e.g. an interesting possible canary in this coal mine is that there’s been a 200% increase in the rate of new apps appearing on Apple’s App Store, but it has not been accompanied by a 200% increase in the rate at which people are buying apps.

If the quality of all apps remains high, but if there is an increase of low quality apps it may not necessarily be great for consumers as it becomes difficult to distinguish which are the good and bad quality apps, making it risky to purchase apps.

Nah the codebase is legacy fucked and I cant be bothered to try and optimize business flows without the fear of other stuff breaking. Claude 100% of the time even thinks we use laravel despite the project being some old lumen codebase, so most of laravels features are not available. It also gets the PHP version we are using wrong 100% of the time.

Have you tried adding this information to claude.md so it knows? I also think your excuse is bad. "The code is legacy fucked so I'll just legacy fuck it some more because I can't be bothered to make an effort"

This is a spicy take, unless the business is willing to face some down time, and I am hired to do exactly what you said, I’d never touch any line of code unless I absolutely have to. Different environments don’t help as much. We tend to obsess over software quality when it’s the least important thing for a business. It’s just a means to an end.

This is what its about, we have multiple ecom shops running 24/7 and cant simply afford downtime or a change of business flow that maybe doesnt affect shop A and B but definitely affects shop C and D...

> Least important thing for a business - Takes weeks or months to get simple features out the door, and when they're out they're buggy as hell and the bugs never get fixed. Sound familiar? > I’d never touch any line of code unless I absolutely have to And this is how legacy code is made. Years of everyone "never touching anything they don't have to" leads to a giant steaming pile of shit. > unless the business is willing to face some down time How does a simple refactor cause downtime? I do this kind of stuff all the time and pretty much never cause any downtime. In the very rare cases that prod downtime does occur it's generally not because of some simple code refactor, and we have it back up in no time by just rolling it back. Unless it's not related to the code at all, in which case it also wasn't a refactor that caused it.

Are you some kind of entitled corporate dev that barely has any influence on the codebase? If I fuck up a whole business goes down as I am the only dev there currently. We cant afford that happening. Also why would I mess with anything claude.md related? I just use the CLI tool. LLM enthusiasts always claim how smart these things are so they should figure it out on their own, you know?

I'm legit annoyed at opus 4.8 at any setting above 4.8. I believe it can be great for vibe coding, but mundane day work? Hell no, I'd rather work with Haiku. It's too slow, checks too many things, it's annoying as hell.

Because companies are betting that this spending will allow them to reduce cost by firing people. Right now the AI LLM PRs we're seeing are just introducing more work for other people, while these so-called builders are looking good with their new dashboards and functionality they're demoing. But you can't talk to them about the flow of the code. You can't ask them for their thinking as to why certain things are. It's not built up from the ground with experience from x people taken into account. It's materialized from nothing, with no foundational separation, and barely any abstractions. No one wants to touch it. The PRs are too large, and the 'authors' of the PRs aren't on call with us. They get all the glory, but do none of the work. It's kinda like designing a house and then sending it to an architect and engineer saying: make this work.

Let's be real. Most of the time you ask an LLM "Why did you do it like this?", it responds with something along the lines of "Oops. My bad. You're right to point this out." You even have a fair chance of getting a response like that when there isn't anything wrong and the question wasn't rhetorical - which perfectly illustrates the level of the genuine understanding LLMs operate at.

A lot of people think they're above average. A lot of them are wrong. A lot of average people are producing gigantic messes. At least previous to this they were gated by their mediocrity.

and have they totally got rid of the average employees? They can blame the models for the production outages already?

That's because of a fundamental misunderstanding of what an LLM is. The only correct answer to "Why did you do it like this?" is that the specific combination of input text and RNG state caused this particular output. There's no reasoning to be had. * EDIT * What's with the downvoting? That's a correct description of what happened. You can't ask an LLM why it did something and expect a coherent response, because there's no thinking chain, and no stored thinking state... At best, you can get a reconstruction of how the context relates to the output (basically a summarization of the context).

Happened to me at least three times the past 14 days. I point out where it made a design decision that causes data loss. «Oops my mistake»

I encounter it constantly with the latest models. Claude is particularly prone to it. > I shouldn’t have said that with confidence > I got ahead of myself there > I overstepped, allow me to correct that It’s wild seeing how often it’s wrong, and I only know it’s wrong because I am an SME or actually reading the sources. Most of my coworkers are not SMEs with what they are asking and do not read the sources. A huge part of my job now is fixing fuck ups and failures resulting from these slop jockeys who have already moved on to slop up the next task.

Literally in the middle of ripping apart a vibe coded mess at work to figure out what's even worth keeping. Not fun :(

What happens if you just keep vibe coding is? Does it whack-a-mole fix one area and break another?

It's so fucking bad. I'm watching a team try to maintain a huge dashboard/control application that interfaces with a large amount of hardware using solely AI workflows. Literally nothing works, all the timers/time counters are different across the pages, constantly commands hardware to do stupid shit, breaks during critical moments/in front of clients. Eventually mgmt had to institute change freezes for high profile events because the team was breaking too much shit all the time. The average C suite dipshit doesn't realize that the performance drops off a cliff once your project is more than some fraction of the context window so they will make pretty dashboards all day long but once you need to cover all the edge cases of a real system it all explodes. AI isn't trained on the type of software style we'll need to create systems using AI, it's trained on how we used to write software. It doesn't reuse code or elegantly structure annoying, it just adds more code until the thing builds and passes some fake tests, even if half of it is functionally dead/unused.

“AI coding is a fad” is not just one big camp of similar-minded people. Different groups have to give up on their pre-existing beliefs in order to be ok with AI coding. Think of people who were very strict with variable names. People who pushed for multiple-levels deep of abstractions for a single API logic that’s not going to be reused. People who believed that coding is craft, rather than just a process to get to the end during work hours. This makes most of these people’s points more-or-less moot. I was in some of those camps, but I’ve seen coding evolve in the last 15 years. So I understand that these priors need to be updated, as most arguments don’t apply to today’s world.

"as most arguments don't apply to today's world" makes me want to roll my eyes so hard at you. The vast majority of problems we had with building complicated systems are all still just sitting there. People are speedrunning relearning things we've known about software engineering for decades. The more things change, the more they stay the same.

So I accept that “nonsense arguments are nonsense”, but with some minor differences of opinion. Naming of things matters insofar as you care as a human to actually conceptualize the system you’re building. You can call all of this stuff minutiae, and on some level I kind of agree, except for the general vibe of _caring about the quality of the stuff you produce_. That is something that still matters whether it “works”. Like, yes you can get an LLM to gen some junk, but _is it any good_ is still something you are in charge of. As far as “boring systems are boring”, I can tell you from experience that I work on a pretty boring system, and AI is not all that meaningful in terms of its impact, and it’s not for a lack of trying. Can it help me create a migration and add an endpoint and such? Sure. But those aren’t the hard problems. They never were. It’s funny that you think the idea of slowing down is such a bad one, but it is another well-established truth. Slow is smooth, and smooth is fast. This notion of break/fixing your way to prosperity by way of 10,000 ill-conceived PRs is a fool’s game.

What's an int vs a float vs a boolean? What's a function? What's a class? What's a variable ? You don't actually need to know the answer to those questions in order to vibe code. That's a lot of priors to update!

yes, but a person who doesn't know any of this stuff is infinitely more productive with ai than someone who isn't when it comes to many things. we've got product folks vibing out prototypes (not shippable but clickable) in our main front end in a few minutes to an hour. This would previously have involved 3 people and several weeks, or a ton of figma and documents to fill in the gaps. This saves weeks to months and lets them really experience the items. Then they hand it off to someone who knows all that stuff who is also using AI and the impl also gets done faster. The PMs are either moving infinitely faster, or at least 30x faster and not blocked constantly by others. basically you're not comparing people who don't know much (tech) with those who do, you're comparing them before and after access to AI.

And, you don't have to vibe code. A competent developer can make great use of AI. I think a developer that can develop the system themselves is the most accelerated user.

> You don't actually need to know the answer to those questions in order to vibe code No, but you do need to know the answer to respond to that 3AM page about prod being down.

1500/Mo per engineer is such a small price considering the base salary of these employees, Maybe Uber knows something we don't (the 5X engineering ROI isn't there for them?). Judging the ROI of an engineer is hard. Adding AI on top of that makes things worse, I think. I've heard AI makes engineers 3X, 5X, 10X and even 100X. If I told my CEO that I was 4X more effective with AI, I am doubtful he would be willing to spend even 1X my salary on tokens. Even though he would be making out in the end. At some point the ROI is pretty much vibes, man.

Because the vibe coded stuff is sometimes great, sometimes it breaks stuff, sometimes it breaks things that we fixed multiple times earlier. The PRs are too large, nobody can review that mess and you better be on call for your deployment. Maybe it will get better, maybe not. I dont know yet.

The massive PRs is something that probably has to end. You can ai generate smaller changes in reviewable PR sizes. It probably even helps the AI code review tools to break the work in to smaller logical chunks too.

If anything the snake is eating it’s own tail because now it’s training on vast amounts of its new slop…dragging down the average bar of quality.

Because writing huge amounts of code is easy for humans too. Agents already proved that they can do it. But are agents able to maintain it? I do not know and unless I know for sure, I am not fully committing to AI generated code. i.e. I am able to write about 1k lines of code of "acceptable" quality per week. Which means in 1 year, there will be about 5Ok LoC. I am pretty sure, that I would have to spent like 60-80% of time to maintain 1st year code and the rest to make new features in the second year so I would have to hire more people and spent time to onboard them to maintain velocity. All of that are rough estimates, probably overoptimistic and way worse in 3rd year. Good luck doing such estimates with code agents. Even worse if you already have huge amounts of legacy code.

yea, and understanding that too is important. the idea you dont need to read code or analysis seems to align with the depwndcy addiction being shoved in thw pipe.

Can you show me an example of a hard task that can't be achieved using light models? When we don't want the model to work on autopilot without reviewing the code at all. Even SOTA models will produce garbage code, if you don't guide them all the time. Hard tasks require a lot of guidance and code reviewing, unless you are creating another throw away project where correctness, maintainability and code understanding does not matter.

Yeah this seems to be a pretty widespread story, from what I've heard as well. The thing about those janky dashboards and spreadsheets though is that somebody understood them and built them with intent to solve a particular problem. Despite the rickety appearance, they're trustworthy tools. A polished single page app might look nicer but it's harder to debug than an excel sheet, and much less transparent in its internal workings--especially if nobody actually wrote it...

Imo its pretty clear that anyone who is taking the issue at least somewhat seriously knows the amount of value they provide is not non-zero. However, the problems are manifold: firstly, toolchains vary wildly, from fancy autocomplete, to engineers chatting with codebases they're unfamiliar with, to people integrating them into devops and infra, to people doing spec driven development, with a thousand philosophies inbetween. Many people suspect that those above them in the ladder are on the cusp of massive failure due to losing track of the code, and many people higher on the ladder think those below them are overly cautious. I hate to be the guy saying "oh it must be somewhere in the middle", but I will say at the very least I like being able to use it to read docs for me, and to synthesize syntax and simple scripts (give me a join that works across these tables and gives me column x, y and z - give me a python script that parses a file like this example and extracts abc data - given this api spec figure out how I can get this data from this endpoint, go) as for building actually complex software, the art of that is not in simply chaining together such scripts. Its the art of using architecture and testing to shape uncertainty, and developing requirements (and extrapolating sensibly from incomplete requirements). I don't think llms are great at this, but they arent terrible either. A lot of the more active users in the space are doing stuff where theyve realised they need more detailed specs, which like, yeah, we knew this already - better defined problems lead to better software.

The point of a refactor is for you to think deeply about the code you are responsibility for, so you can make it better (faster, easier to work on, more tests, whatever). You’ve gotten a result, but without the work that made you valuable, while deskilling yourself. It’s a lose/lose situation for…I would say anyone employed as an engineer or programmer. I’m not taking responsible for AI output, the same way I won’t try to fix auto-generated code: because you just regenerate it. The only person that wins here is the person who can pay you less because they don’t need you, they just need another “types computer guy”.

Sure they’re fine at that sort of rote find/replace job as long as it’s relatively straightforward. But it only really works if you do the hard parts yourself then tell the agent to go and do the rote part. Even then I’ve had it turn to slop more often than not as the agent has to start contorting the code into weird shapes to try and finish the job. It’ll never stop and be like “hey maybe this was a bad idea, let’s try something else”. And by the time you get to review it, you’ve spent 20 bucks on something that needs to be thrown away.

That's not at all negative about LLMs, just negative about his own usage of LLMs. He's still very heavily and unrealistically (unless he has very poor coding standards and skills, which I won't rule out) praising LLMs in the sentences you've quoted. Note that it's not surprising that he finds his own usage (described in the quote) negative, since his real job is as a blogger, not anything else.

I'm curious how much of the usage comes from vibe coding vs using agents/harnesses in internal tooling

I think there's too much variance between what model you're using and how much you turn your brain off. If I just paste a ticket number into 4.8xHigh its going to use a lot more tokens than if I read the ticket, tell Sonnet what it needs to do, make my commit, run unit tests myself, etc.

Reading the headline Oh that's actually really economical! I wonder if they're doing a lot on locally running models or managing a shared context or knowledge-base in some clever way, maybe just encouraging employees to be efficient and mindful. ... > each employee ... > per AI coding tool ... > I noted that my own token usage comes to about $1,000/month against each of Anthropic and OpenAI What on this godforsaken earth are all you rich idiots doing???

I still have never hit a ceiling with my Claude Max $100 account, much less the Max $200 account. I'm not burning tokens needlessly, nor running it all day, but I do use CC almost daily. What are these devs doing that they are burning more than $1500 in tokens a month? Maybe it's just me, but I still find that I really have to "shepherd" the AI and work with it to get the results I want. And I read every line of code added and challenge the model's logic. So that limits my token burning. Maybe these people are just "vibe-coding" without really checking the results?

I would not be surprised if they have engineers vibecoding 2-3 projects each simultaneously, nonstop, on largely un-moderated review-suggest-iterate-test feedback loops. All the code gets summarized and fed into their manager's agent contexts, probably duplicated several times across levels and departments, with some generated back-and-forth emails pinging around the org chart, eventually generating 2-3 long-winded reports that nobody will read chock full of generated visualizations that can all get consolidated into a generated slide deck that they'll show (maybe, at some point) to a handful of humans with more money than a human brain can conceptualize to demonstrate all of the innovation they're doing. I am increasingly convinced that many of these companies are dead trees whose only function is to burn money lest it fall into the hands of the peasantry.

just don't care about the output. Produce more. Don't check the results.

I have strong conviction that companies will now choose tech stack/programming languages based on 'tokenomics'. I am vibe coding using Clojure, a language I can read but cannot write and I never hit the usage limits even when using the latest model on Claude. I have similar experience with F#, which is a bit more verbose than clojure but absolutely beats every OOP language, Python, Typescript etc. The reason, I use F# & Clojure is they hit JVM and CLR, two popular enterprise stacks. In my not so humble opinion Lisp(Clojure) still remains the language of AI.