Vibe coding and agentic engineering are getting closer than I'd like

While those who are hands on is realising the limits and issues with vibe/context engineering/agentic engeering/buzz-word-of-the-week the businesses and pushing hard on the buzz words. It’s high time we start looking at ways to live with the new reality and figure out ways to ensure software reliability.

The disconnect for AI is that it is a jagged frontier and it only really shines when one of its jagged frontiers extends counter to one of your valleys.

If you've been writing Perl for 30 years, you might not want to learn JavaScript just to make a little fun idea in your head to show your wife. Vibe code that shit man. Who cares? Your wife does not care about LOC or those internal design decisions you made.

If you're trying to learn something new like an algorithm, protocol, or API write that shit by hand. You learn by doing, and when you know how the thing works and have that mental context, you will always be faster than an AI. Also, when did we stop liking to learn? Why is it a bad thing to know all the ins and outs of a programming language? To write and make all the decisions yourself? That shit is fun. I don't care if you disagree.

If you're at work and they really care about getting something out of the door, do whatever you think is best. If you just wanna ship vibed code and review PRs all day, all the power to you. If you wanna write it by hand, and use AI like a scalpel to write up boiler plate, review code, do PR audits, etc... go for it!

A hammer is a really great tool that has thousands of purpose-designed uses. I still prefer my key to get into my car. It's all tools, you are a person.

A lot of this stuff if coming top-down from people who do not have the experience you do. Wouldn't a smart employee use their expertise to advise the organization? If you work at a company where that would not be okay, maybe it's time to start looking for another firm.

Vibe Coding (and LLMs) did not create undisciplined engineering organizations or engineers. They exposed and accelerated them.

Plenty of engineers have loose (or no!) standards and practices over how they write coee. Similarly, plenty of engineering teams have weak and loose standards over how code gets pushed to production. This concept isn't new, it's just a lot easier for individuals and teams who have never really adhered to any sort of standards in their SDLC to produce a lot more code and flesh out ideas.

> I know full well that if you ask Claude Code to build a JSON API endpoint that runs a SQL query and outputs the results as JSON, it’s just going to do it right. It’s not going to mess that up. You have it add automated tests, you have it add documentation, you know it’s going to be good.

I feel like this is just not true. An JSON API endpoint also needs several decisions made.

- How should the endpoint be named

- What options do I offer

- How are the properties named

- How do I verify the response

- How do I handle errors

- What parts are common in the codebase and should be re-used.

- How will it potentially be changed in the future.

- How is the query running, is the query optimized.

…

If I know the answer to all these questions, wiring it together takes me LESS time than passing it to Claude Code.

If I don’t know the answer the fastest way to find the answer is to start writing the code.

Additionally, whilst writing it I usually realize additional edge cases, optimizations, better logging, observability and what else.

The author clearly stated the context for this quote is production code.

I don’t see any benefits in passing it to Claude Code. It’s not that I need 1000s of JSON API endpoints.

Perhaps I've missed a few weeks worth of progress, but I don't think that AIs have become more trustworthy, the errors are just more subtle.

If the code doesn't compile, that's easy to spot. If the code compiles but doesn't work, that's still somewhat easy to spot.

If the code compiles and works, but it does the wrong thing in some edge case, or has a security vulnerability, or introduces tech debt or dubious architectural decisions, that's harder to spot but doesn't reduce the review burden whatsoever.

If anything, "truthy" code is more mentally taxing to review than just obviously bad code.

> If you can go from producing 200 lines of code a day to 2,000 lines of code a day, what else breaks? The entire software development lifecycle was, it turns out, designed around the idea that it takes a day to produce a few hundred lines of code. And now it doesn’t.

It is so embarrassing that LOC is being used as a metric for engineering output.

Have you noticed that the coding agents get really close to the solution on the first one shot and then require tons of work to get that last 10% or 5%?

If we shift the paradigm of how we approach a coding problem, the coding agents can close that gap. Ten years ago every 10 or 15 minutes I would stop coding and start refactoring, testing, and analyzing making sure everything is perfect before proceeding because a bug will corrupt any downstream code. The coding agents don't and can't do this. They keep that bug or malformed architecture as they continue.

The instinct is to get the coding agents to stop at these points. However, that is impossible for several reasons. Instead, because it is very cheap, we should find the first place the agent made a mistake and update the prompt. Instead of fixing it, delete all the code (because it is very cheap), and run from the top. Continue this iteration process until the prompt yields the perfect code.

Ah, but you say, that is a lot of work done by a human! That is the whole point. The humans are still needed. The process using the tool like this yields 10x speed at writing code.

For me the distinction is the quality and rigor of your pipeline.

Vibe coding: one shot or few shot, smoke test the output, use it until it breaks (or doesn't). Ideal for lightweight PoC and low stakes individual, family or small team apps.

Agentic engineering: - You care about a larger subset of concerns such as functional correctness, performance, infrastructure, resilience/availability, scalability and maintainability. - You have a multi-step pipeline for managing the flow of work - Stages might be project intake, project selection, project specification, epic decomposition, d=story decomposition, coding, documentation and deployment. - Each stage will have some combination of deterministic quality gates (tests must pass, performance must hit a benchmark) and adversarial reviews (business value of proposed project, comprehensiveness of spec, elegance of code, rigor and simplicity of ubiquitous language, etc)

And it's a slider. Sometimes I throw a ticket into my system because I don't want to have to do an interview and burn tokens on three rounds of adversarial reviews, estimating potential value and then detailed specification and adversarial reviews just to ship a feature.

Yup, the normalization of deviance here is a real thing. I still review all the code the LLM generates (well, really, I have it generate very little code: I use it more for planning, design, rubber-ducking, and helping track down the causes of bugs), but as time goes on without obvious errors, it gets more and more tempting to assume the code is going to be fine, and not look at it too closely.

But resisting that impulse is just another part of being a professional. If your standards involve a certain level of test coverage, but your tests haven't flagged any issues in a long time, you might be tempted to write fewer tests as you continue to write more code. Being a professional means not giving in to that temptation. Keep to your quality standards.

Sure, standards are ultimately somewhat arbitrary, and experience can and should cause you to re-evaluate your standards sometimes to see if they need tweaking. But that should be done dispassionately, not in the middle of rushing to complete a task.

And hell, maybe someday the agents will get so good that our standards suggest that vibe coding is ok, and should be the norm. But you're still the one who's going to be responsible when something breaks.

"I want professionally managed software companies to use AI coding assistance to make more/better/cheaper software products that they sell to me for money.” (Simon Willison herein quotes Matthew Yglesias) - this is such a naive and sloppy take. What do you want? "better software"? not going to happen. "cheaper software"? not going to happen either. "more software"? for sure, but is it really what you want?

If I hire a plumber it's certainly not cheaper than doing it myself but when I am paying money I want to make sure it is better quality than what I am vibe plumbing myself.

Here's for the AI supremacists:

Let's assume AI is 10x perfect than humnas in accuracy and produces 10x less bugs and increases the speed by 1000x compared to a very capable software engineer.

Now imagine this: A car travels at a road that has 10x more bumps but it is traveling 1000x slower pace so even though there are 10x bumps, your ride will feel less bumpy because you're encountering them at far lower pace.

Now imagine a road that has 10x less bumps on the road but you're traveling at 1000x the speed. Your ride would be lot more bumpy.

That's the agentic coding for you. Your ride would be a lot more painful. There's lots of denial around that but as time progresses it'll be very hard to deny.

Lastly - vibe coding is honest but agentic coding is snake oil [0] and these arguments about having harnesses that have dozens of memory, agent and skill files with rules sprinkled in them pages and pages of them is absolutely wrong as well. Such paradigm assumes that LLMs are perfect reliable super accurate rule followers and only problem as industry that we have is not being able to specify enough rules clearly enough.

Such a belief could only be held by someone who hasn't worked with LLMs long enough or is a totally non technical person not knowledgeable enough to know how LLMs work but holding on to such wrong belief system by highly technical community is highly regrettable.

[0]. https://news.ycombinator.com/item?id=48018018

>> The entire software development lifecycle was, it turns out, designed around the idea that it takes a day to produce a few hundred lines of code.

Yeah. I'm not sure how other people work, but I almost never need to write formal tests because I essentially test locally as I write, one method at a time, and at that moment I have a complete mental map of everything that can potentially go wrong with a piece of code. I write and test constantly in tandem. I can write a test afterwards to prove what I already know, but I already know it. This is time consuming, anal, and obsessive-compulsive, and luckily that kind of work perfectly suits my personality. The end result is perfect before I commit it.

It is a lot of fun asking LLMs to write code around my code. Make 10 charts with chartjs in an html page that show something and put it behind a reverse proxy so the client can see it. Wow. Spot on, would've taken me an hour. I can even rely on Claude to somewhat honestly reason about things in personal projects.

But knowing every implementation decision makes a huge difference when anything real is at stake. "Guilt" wouldn't begin to describe the sense I'd have id my software did something because of a piece of code I hadn't personally reviewed and fully understood, at which point I probably should have just written it myself.

> The entire software development lifecycle was, it turns out, designed around the idea that it takes a day to produce a few hundred lines of code. And now it doesn’t.

No, it was never designed around that. All methodologies of software dev don't focus too much on writing the code, but on everything else: requirement definition, quality, maintenance, speed of integrating feature, scaling the work, ...

Personally with 20 years of experience, I never seen a single company were writing the code was a bottleneck

When I was in grad school I graded homework for first year math classes, and the thing about math homework is that the perfect homework takes almost no time to grade.

It's the bad, semi-coherent submissions that eat up your time, because you do want to award some points and tell students where they went wrong. It's the Anna Karenina principle applied to math.

Code review is the same thing. If you're sure Claude wrote your endpoint right, why not review it anyway? It's going to take you two minutes, and you're not going to wonder whether this time it missed a nuance.

I think all coding will become vibe coding, but it will be no less an engineering discipline.

Note: I still review pretty much every line of code that I own, regardless of who generates it, and I see the problems with agents very clearly... but I can also see the trends.

My take: Instead of crafting code, engineering will shift to crafting bespoke, comprehensive validation mechanisms for the results of the agents' work such that it is technically (maybe even mathematically) provable as far as possible, and any non-provable validations can be reviewed quickly by a human. I would also bet the review mechanisms would be primarily visually, because that is the highest bandwidth input available to us.

By comprehensive validations I don't mean just tests, but multiple overlapping, interlocking levels of tests and metrics. Like, I don't just have an E2E test for the UI, I have an overlapping test for expected changes in the backend DB. And in some cases I generate so many test cases that I don't check for individual rows, I look at the distribution of data before and after the test. I have very few unit tests, but I do have performance tests! I color-code some validation results so that if something breaks I instantly know what it may be.

All of this is overkill to do manually but is a breeze with agents, and over time really enables moving fast without breaking things. I also notice I have to add very few new validations for new code changes these days, so once the upfront cost is paid, the dividends roll in for a long time.

Now, I had to think deeply about the most effective set of technical constraints that give me the most confidence while accounting for the foibles of the LLMs. And all of this is specific to my projects, not much can be generalized other than high-level principles like "multiple interlocking tests." Each project will need its own custom validation (note: not just "test") suites which are very specific to its architecture and technical details.

So this is still engineering, but it will be vibe coding in the sense that we almost never look at the code, we just look at the results.

I want to agree, I do. But this point is plainly wrong in my observations:

> The enterprise version of that is I don’t want a CRM unless at least two other giant enterprises have successfully used that CRM for six months. [...] You want solutions that are proven to work before you take a risk on them.

Perhaps not for every category of software and every company. But in practice, any SaaS app that is just CRUD with some business logic + workflows is, imo, absolutely vulnerable to losing customers because people within their customers' orgs vibe coded a replacement.

They are perhaps even more at risk because would-be new customers don't ever even bother searching to find them as an option because they just vibe code a competitor in-house.

The vulnerability lies primarily in the fact that most of these SaaS apps were talking about are _wrong_ to some meaningful degree. They don't fully fit how your company works, and they never did. There is something about them that you are forced to work around in some way. This is true because it is impossible to build a universally perfect product, to perfectly fit it to every business requirement of every user in every company.

But now it is relatively cheap to build the perfect version for your company in-house. Or maybe even just for YOU.

I think medium/long-term this will mean a redistribution of technical talent from SaaS companies to industry companies. Instead of paying millions for SaaS subscriptions, industry companies will spend fewer millions building precisely what they need in-house with the help of AI. Not every SaaS and not every company, but I already see this happening at my company right now.

What an excellent article by a smart, humble, still-learning person!

Favorite quote:" There are a whole bunch of reasons I’m not scared that my career as a software engineer is over now that computers can write their own code, partly because these things are amplifiers of existing experience. If you know what you’re doing, you can run so much faster with them. [...]

I’m constantly reminded as I work with these tools how hard the thing that we do is. Producing software is a ferociously difficult thing to do. And you could give me all of the AI tools in the world and what we’re trying to achieve here is still really difficult. [...]"

The scary part is that codebases are getting layers of AI complexity, that it's going to cost $$$ to have the latest model decipher and make changes as no human can understand the code anymore.

Pretty soon there is no code reuse and we're burning money reinventing the wheel over and over.

> It’s not just the downstream stuff, it’s the upstream stuff as well. I saw a great talk by Jenny Wen, who’s the design leader at Anthropic, where she said we have all of these design processes that are based around the idea that you need to get the design right—because if you hand it off to the engineers and they spend three months building the wrong thing, that’s catastrophic.

This is spot on. I think the tooling is evolving so much particularly on the design side that its not worth the "translation cost" to stay (or even be) on the Figma side anymore.

Claude often does things in more detail, and even better, than I would, in the first pass. But I don't understand how anybody stands comments generated by an LLM?

It's seriously the thing that worries (and bothers) me the most. I almost never let unedited LLM comments pass. At a minimum.

Most of the time, I use my own vibe-coded tool to run multiple GitHub-PR-review-style reviews, and send them off to the agent to make the code look and work fine.

It also struggles with doing things the idiomatic way for huge codebases, or sometimes it's just plain wrong about why something works, even if it gets it right.

And I say this despite the fact that I don't really write much code by hand anymore, only the important ones (if even!) or the interesting ones.

Also, don't even get me started on AI-generated READMEs... I use Claude to refine my Markdown or automatically handle dark/light-mode, but I try to write everything myself, because I can't stand what it generates.

This is my workflow which I find very productive with Agentic AI.

Disclaimer: I'm doing a CAD-like engineering desktop app, and I'm using VS 2026 Copilot, so YMMV.

When I get a Jira ticket, I will first diagnose the problem, and then ask AI to write a test case for it that will reproduce the problem, with guidance on what/how to do the test case (you will be surprised to know how many geometry, seemingly visual problems can be unit tested), and if necessary I provide clues (like which files to read, etc.) for AI to look at, and ask AI to just go and fix the test.

Often AI can do that; AI can make the test pass and make sure that adjacent tests also pass. If in doubt, I will check the output reasoning. I then verify that the fix is done properly via visual inspection (remember, this is a desktop app), and I ask for clarification if needed.

Then at night I'll let my automated test suites run... and oops! Regression found! Who broke it? AI or human? Who cares. I just tell AI that between these times one of the commits must have broken the code — can you please fix it for me? And AI can do that.

This works for small or medium feature implementation, trival bugfixes, or even annoying geometrical problems that require me to dig out the needle in the haystack. So the productivity gain is very real. But I haven't tried it on feature that requires weeks or months for implementation, maybe I should try it next time.

It's hard to describe the feeling. It's just that the AI is working like a very capable (junior?) programmer; both might not have full domain knowledge, but with strong test suites and senior guidance, both can go very far. And of course AI is cheaper and a lot more effective.

The real paradigm shift is not here yet, but not very far away. I'm talking about the single unified codebase. Agents building a unique codebase for all your software needs.

Because most of the complexity in software comes from interfacing with external components, when you don't need to adapt to this you can write simpler and better code.

Rather than relying on an external library, you just write your own and have full control and can do quality control.

Linux kernel is 30 000 000 LOC. At 100 tokens /s, let's say 1 LOC per second produced for a single 4090 GPU, in one year of continuous running 3600 * 24 * 365 = 31 536 000 everyone can have its own OS.

It's the "Apps" story all over again : there are millions of apps, but the average user only have 100 max and use 10 daily at most.

Standardize data and services and you don't need that much software.

What will most likely happen is one company with a few millions GPUs will rewrite a complete software ecosystem, and people will just use this and stop doing any software because anything can be produced on the fly. Then all compute can be spent on consistent quality.

>If you can go from producing 200 lines of code a day to 2,000 lines of code a day, what else breaks? The entire software development lifecycle was, it turns out, designed around the idea that it takes a day to produce a few hundred lines of code. And now it doesn’t.

How is producing more lines of code any good? How does quality assurance work with immeasurable code bloat? I want good software not slopware with 2000 different features. A good product does few things, but does these really well. There is no need to constantly add lines of code to a working product.

As a web developer, I feel like this take is wildly optimistic. My remaining qualifications that still provide some sort value are providing historical/business/architectural context to the agent and testing the agent's output. And that's only because 1) it's not all written down in Markdown and 2) the agent is massively nerfed by costs and Anthropic. The thing in the middle where I get a coffee and write code in a variety of languages, then pop open a debugger has been fully obsoleted.

Its just economy 101.

People have been running crappy code commercially for over half a century now. Not many companies successfully differentiate by running good code - it usually does not matter to the end consumer, other things are much more important. So now companies will pay less for code, and maybe it is a bit worse (though I personally can't believe AI can do worse than corporate software developers on average). Hobbyists will remain hobbyists, and precious few will be lucky enough to have someone pay them to handcraft stuff. Exactly what happened to woodworkers and other craftsmen.

In my own experience, good engineering practices are still not easy to achieve. As a software engineer with three years of experience, I've been doing solo dev for the past few months. Currently, there is still a lot of the harness to set up manually.

> The thing that really helps me is thinking back to when I’ve worked at larger organizations where I’ve been an engineering manager. Other teams are building software that my team depends on.

> If another team hands over something and says, “hey, this is the image resize service, here’s how to use it to resize your images”... I’m not going to go and read every line of code that they wrote.

The distance of accountability of the output from its producer is an important metric. Who will be held accountable for which output: that's important to maintain and not feel the "guilt".

So, organizations would need to focus on better and more granular building incentives and punishment mechanisms for large-scale software projects.

There are techniques for improving our confidence in our software: unit testing, integration testing, fuzz testing, property-based testing, static analysis, model checking, theorem proving, formal methods, etc. The LLM is not only a tool for generating lines of code. It can also generate lines of testing. The goal is that the tests are easier to audit by the humans than the code.

I guess it all depends on what you use it for.

I work on database optimizers and other database related stuff, and I can assure Claude Code - with all the highest settings - does make mistakes. It will generate a test that does not actually test what it "thinks" it tests. It will confidently break stuff.

Do not get me wrong. It is still awesome! It takes much of grunt work off me. It can game out designs decisions even when that needs to refactor a lot of code. If you point out a mistake more often than not it can fix it itself.

It's just for a critical project I would never ship it without understanding every line of code - with the exception perhaps of some of the test code. Maybe in a year or two that will be different.

Repeat after me: most software spends the majority of its lifetime in the maintenance phase.

Repeat after me: it follows that most of the money the software makes occurs during the maintenance phase.

Repeat after me: our industry still does not understand this after almost 100 years of being in existence.

Alan Kay was 100% right when he said that the computer revolution hasn't occurred yet. For all of our current advancements all tools are more or less in the Stone Age.

My great hope is that AI will actually accelerate us to a point where the existing paradigm fully breaks beyond healing and we can finally do something new, different, and better.

So for now - squeee! - put a jetpack on your SDLC with AI and go to town!!! Move fast and break things (like, for real).

2 days ago, we updated a stripe library which broke everything. With AI, I was able to one shot wrapping all of the calls into a shared service, patched the broken api contract across the entire app and got our signup and payment flows working again. solid day and a half of work. this would have taken a days of back and forth debugging previously. AI is not a panacea for everything but its doign valuable work right now.

I agree somewhat, but I do still think there is a decently sized separation between true vibe coding (the typical "make me an app...fix this bug") and actual AI assisted development. I personally think that if you are a dev and you simply trust the AI's output, that is still vibe coding.

I am not a developer and have very basic code knowledge. I recently built a small and lightweight Docker container using Codex 5.5/5.4 that ingests logs with rsyslog and has a nice web UI and an organized log storage structure. I did not write any code manually.

Even without writing code, I still had to use common sense in order to get it in a place I was happy with. If i truly knew nothing, the AI would have made some very poor decisions. Examples: it would have kept everything in main.go, it would have hardcoded the timezone, the settings were all hardcoded in the Go code, the crash handling was non existent, and a missing config would have prevented start. And that is on a ~3000 line app. I cannot imagine unleashing an AI on a large, complex. codebase without some decent knowledge and reviewing.

The "has someone actually used it" signal is the new code review. Tests, docs, commit count all reproducibl in 30 minutes. Daily usage for 2 weeks isn't. That's the only proof of work that survived the agent era.

This is a timely observation and feels right to me. I needed to get a relatively simple batch download -> transform -> api endpoint stood up. I wrote a fairly detailed prompt but left a lot of implementation details out, including data sources.

Opus 4.7 built it about 90% the same way I would, but had way more convenience methods and step-validations included.

It's great, and really frees me up to think about harder problems.

One-shot "vibe coding" is generally a mistake.

But using an agentic LLM to complete boilerplate is attractive simply because we've created a mountain of accidental and intentional complexity in building software. It's more of a regression to the mean of going back to the cognitive load we had when we simply built desktop applications.

From the podcast episode they talk about the idea of using an LLM for training by disallowing the model to write code. I've been experimenting with exactly that in conjunction with a proof checker (Agda) to help me learn some cubical type theory and category theory.

I find the LLM as interactive tutor reviewing my work in a proof checker to be a really killer combo.

It's already the case that you get much better results out of LLMs by forcing agents using them to go through additional layers of planning, design & review.

The future is going to dynamically budget and route different parts of the SLDC through different models and subagents running on the cloud. Over time, more and more of that process will be owned by robots and a level of economic thinking will be incorporated into what is thought of today as "software engineering." At some point vibe coding _is_ coding and we're maybe closer to that point than popularly believed.

The gap between "vibe coding" and "agentic engineering" is the same gap between asking someone to do a task and being able to prove they did it correctly. One is vibes. The other is accountability. We keep building more powerful agents without building the audit infrastructure to verify what they actually did.

The more I use AI, the more I find it’s great for anything trivial and uninspired. Need help with some predictable glue code? AI. Need help with something insightful and new to the world? Not AI. Need help with an important task that’s been done a 1000 times? AI with scrutiny. Need to invent something new to the world and core to your business? Probably not AI.

We still have not the right sandbox and PR abstractions to make the merge of the two complete. Imagine merging a PR and knowing exactly this code cannot ever possibly reach the internet and it can only receive and send specific shapes of api requests from these specific services, it has well defined resource limits and you have specific optimal UI to review these constraints. I can imagine to not review a bigger number of PRs in that reality.

Used to check every line for my project. Now i just check the tricky parts still don't know if that's ok or just lazy?

About two years ago I was using the term "agentic engineer" to describe someone who builds AI agents - not a vibe coder.

Agentic Engineer does not make much sense to be applied to a developer.

It is weird and confusing to call a web designer that uses AI assisted coding tools "agentic engineer".

It makes sense that they merge over time; it's a mark of the progress being made. The ultimate end is to make them indistinguishable, where the purely vibe coded app will have the quality of the app that has been well engineered over significant time thanks to good user feedback.

The distinction between 'vibe coding' and 'agentic engineering' is important. In my experience, the key difference is whether you're reviewing and understanding the code the agent produces. When I use coding agents for non-trivial tasks, I always review the diff before committing — that's the engineering part. The danger is when people skip that step and just trust the output.

One thing I've started appreciating with LLM-assisted workflows is how important fixed evaluation protocols are.

Without pre-defined definitions and locked procedures, it's extremely easy to mistake iterative adaptation for genuine signal.

Strong agree. Most orgs will stay tangled in the mess they hand-coded over the years, a few greenfield teams will pull ahead, but until some LLM-fuelled startup displaces a strong incumbent I'm skeptical that we're on the cusp of anything other than a K-shaped transition. I see already low quality software and orgs getting flushed to make room for some new ideas now that the barrier to entry is slightly lower (but far from free). I just wish the transition was done with more humanity.

Given rapidly decelerating quality of, at least, claude code output, the agentic coding use may decrease. It is insane how bad the results of background agents are now: constant hallucinations, nonsensical outputs.

I think I'm just too opinionated to go there. If I see something that works fine, but isn't the way I'd do it, it doesn't matter if a human or an LLM wrote it I'm still in there making it match my vision.

> my disturbing realization that vibe coding and agentic engineering have started to converge in my own work.

>I firmly staked out my belief that “vibe coding” is a very different beast from responsible use of AI to write code, which I’ve since started to call agentic engineering

Disturbing? Really? I admit I don't do agentic and am going only by vibes, but for me agentic engineering is basically vibe coding in a automated loop with some ornamentals. They both stem from the same LLM root and positioning them as significantly different is weird and unconvincing to me. There may be a merit to this article (I gave up after few sentences), but I reject this specific premise.

Software engineering is software engineering.

An ace software engineer is not an ace because of tooling.

It's not the plane, it's the pilot, or something like that.

The current state of the technology is that you must read at least some of the code, but everyone keeps shipping tools that are focussed on churning out more and more stuff without giving you any affordances to really understand the output.

Claude Code in particular seems really uninterested in this aspect of the problem and I've stopped using entirely because of this.

Still thinking about LLM's

Totally agree. The sales pitch is that anyone can use this stuff, but good output is only obtained via thorough understanding.

Never really bought that there was a clean distinction.

To me it’s a spectrum with varying levels of structure provided, review etc.

Basically oneshot vibes on one side, fully hand coded on other.

Correct me if I’m wrong Simon, but weren’t you highly optimistic about llm’s and agentic-use of them?

I believe this is a common fault of not being able to zoom out and look at what trade offs are being made. There’s always trade-offs, the question is whether you can define them and then do the analysis to determine whether the result leaves you in a net benefit state.

No offense, but if feels to me the author writes this piece to convince himself. I am afraid he is right. But the bottom line is the same: vibe coding, agenting engineering, everything AI-related comes for our jobs.

the discourse around "code quality" has always attracted the least nuanced minds, ones who see the world and the phenomenon of life as nothing but territory to be divided up by the latest buzzwords. the worst ones insist that we narrow the discussion even further, to focus on the conflicts between these buzzwords. whenever i have to sit through such discussions, i try to meditate on the irony of mother nature weaving the most functionally brutal, ruthlessly redundant poetry that is the genetic code, only for the resulting creatures to deny themselves the power of the principles inherent in their own construction.

An AI cannot be held accountable to mistakes, so an AI should not be doing your job for you. End of discussion.

As agents get better at code we trust them to produce more of it. There are still bugs to find, but the haystack gets bigger.

So the number of bugs to find remains constant but the amount of code to review scales with the capability of the agent.

I agree to some extent. I think that small aps, dashboards, service wrappers etc. you can vibe code.

But building software still requires domain knowledge, understanding data structures, architecture, which services to use. We probably have 2-5 years before thats fully automated.

I am experimenting with writing en entire TypeScript compiler[1] with AI assistant. I've spent 4 months on it already. It might not be successful at the end of the day but my thinking is that if LLMs are going to write a lot of the code I better learn how this can and can not work. I've learned a lot from this project already. I think we're still in charge of design and big ideas even if all of the code is written by AI

[1] https://github.com/mohsen1/tsz

I can't really say I agree with this, although I also hate the phrase "agentic engineering".

I'm working on a licensing system for a product I'm building. I've used Claude a little bit to help out with it, but it's also made a lot of very dumb decisions that would have large (security!) consequences if I didn't catch them. And a lot of them are braindead things, like I asked it to create a configurable limit on a certain resource for the trial version of the application. When I said configurable, I mostly meant: put the number in a constant so I can update it later. What Claude thought I asked was "make it so the user can modify the limits of the trial version in the settings panel" (which defeats the entire purpose of a free trial!). Another thing it messed up recently is I was setting up email-magic-link authentication. It defaulted to creating an account for anyone that typed in an email, which could allow a bad actor to both spam people with login requests (probably getting me kicked off Resend) or creating a lot of bogus accounts.

These things do not think. You cannnot outsource your thinking to them.

The problem with vibe coding closer is that the agentic makes a very plasticy samey feel unless you work with something that makes it unique or can pass a template through it.

"But I’m not reviewing that code. And now I’ve got that feeling of guilt: if I haven’t reviewed the code, is it really responsible for me to use this in production?"

"I know full well that if you ask Claude Code to build a JSON API endpoint that runs a SQL query and outputs the results as JSON, it’s just going to do it right. It’s not going to mess that up. You have it add automated tests, you have it add documentation, you know it’s going to be good."

This really is Wordpress and early PHP all over again, but it's the seasoned folks rather than the amateurs that buy into it.

I believe these tools will be refined and locked down and eventually turn into RAD stuff used by certified enterprise consultants, much like SAP and Salesforce and IBM solutions and so on. From this I come to the conclusion that it is not a good idea to become dependent on them at this stage, which is corroborated by the pecuniary expense as well as excruciatingly fast change in available products.

Why is it one or the other and not one THEN the other?

What the F is "agentic" really?

Agentic engineering? That reads to me a little like amateur oncologist. How are you defining engineering?

Can agentic engineers adhere to a similar code of ethics that a professional engineer is sworn to uphold?

https://www.nspe.org/career-growth/nspe-code-ethics-engineer...

I mean... yeah? Isn't it obvious that they're essentially the same thing, but one thinks they're in a higher class than the other?

Fast feedback loops and delegating tasks to sub-agents have been pretty common for vibers since well before they were canonicalized by agenteers. Same thing, different day, hardly even any difference in quality: they evolve together, though vibe tends to lead and agents follow and refine... which vibers then use too.

If you think of vibe coders as agentic alpha testers it makes a lot more sense.

> But I’m not reviewing that code (...)

That's the spirit, I always say - _others_ will deal with AI slop during code review. Eventually they will get tired and start 'reviewing' this AI stuff with AI - so it's a win win. Right?

I think this is what people mean when they say LLMs are a higher level abstraction. We still need to consider edge cases and have tests. We still to sweat the architecture and understand how the pieces fit together and have a mental map of the codebase. But within each bottom node of that architecture we don't sweat the details. Anything obvious gets caught right away. Most subtle/interaction-based issues occur at the architecture level. Anything that bypasses those filters is a weird bug that is no worse or different from a normal bug fixes - an edge case that was hit in a real world scenario that gets flagged by a user or a logged as an error.

There are certain codebases and pieces of code we definitely want every line to be reasoned and understood. But like his API endpoint example, no reason to fuss with the boilerplate.

This has definitely been my shift over the past few months, and the advantage is I can spend much more time and energy on getting the code architecture just right, which automatically prevents most of the subtle bugs that has people wringing their hands. The new bar is architecting code to be defined as well as an API endpoint->service structure so you can rely on LLMs to paint by numbers for new features/logic.

> I’m starting to treat the agents in the same way. And it still feels uncomfortable, because human beings are accountable for what they do. A team can build a reputation. I can say “I trust that team over there. They built good software in the past. They’re not going to build something rubbish because that affects their professional reputations.”

The most important part and why slop isn't the same as a code written by someone else. The model doesn't care, it just produces whatever it is asked to produce. It doesn't have pride, it doesn't have ego, it doesn't artisanal qualities, it doesn't have ownership.

Every time I do deep work, and think of solutions to a complex problem. I always have the opportunity to ask claude to implement a sub-par AI slop solution.

Do this enough times, and I will have forgotten how to think.

People in the future are going to wonder what the hell we were thinking, when 30 years down the line everything is a hot mess of billions of lines of code generated by LLMs that no human has read almost any of it and is no longer possible for anyone to maintain neither with nor without LLMs. And the LLM generated garbage will have drowned out all of the good quality code that ever existed and no one will be able to find even human generated code anymore on the internet.

Makes me want to just give up programming forever and never use a computer again.

I feel like an outlier in all of this. But isn't this just more AI slop? How is this different from text generation or image generation?

Like many people I have used AI to generate crap I really don't care about. I need an image. Generate something like, whatever. Great hey a good looking image! No that's done I can do something I find more interesting to do.

But it's slop. The image does not fit the context. Its just off. And you can tell that no one really cared.

This isn't good.

> It used to be if you found a GitHub repository with a hundred commits and a good readme and automated tests and stuff, you could be pretty sure that the person writing that had put a lot of care and attention into that project.

I think this highlights a problem that has always existed under the surface, but it's being brought into the light by proliferation of vibeslop and openclaw and their ilk. Even in the beforetimes you could craft a 100.0% pure, correct looking github repo that had never stood the test of production. Even if you had a test suite that covers every branch and every instruction, without putting the code in production you aren't going to uncover all the things your test suite didn't--performance issues, security issues, unexpected user behavior, etc.

As an observer looking at this repo, I have no way to tell. It's got hundreds of tests, hundreds of commits, dozens of stars... how am I to know nobody has ever actually used it for anything?

I don't know how to solve this problem, but it seems like there's a pretty obvious tooling gap here. A very similar problem is something like "contributor reputation", i.e. the plague of drive-by AI generated PRs from people (or openclaws) you've never seen before. Stars and number of commits aren't good enough, we need more.

I still don’t get what agentic engineering is. Isn’t it all just asking the same LLM what you want it to do?

huh. i honestly never thought they were all that different. didn't the same guy coin them both to refer to the same thing?

Reminder, cybersecurity will be huge in following years.

Companies are shipping things and nobody understands what they're shipping.

> I know full well that if you ask Claude Code to build a JSON API endpoint that runs a SQL query and outputs the results as JSON, it’s just going to do it right. It’s not going to mess that up.

> Claude Code does not have a professional reputation!

how come?

I agree, I'm actually generating just over of 20,000 lines of code each day at my company. Part of that was the mandate and leaderboards around token usage, but also they started using pull requests as an explicit metric. What I do is usually pull around 5 or so tickets at once, spin up 5 different agents on their own branch, have them work until completion, and then spin up two more agents to handle the merge request.

I'm not checking the code since the code doesn't really matter anymore anyways - I just have the agent write passing tests for the changes or additions I make, and so even if something breaks I can just point to the tests.

Some days, the tickets are completed much faster than I expect and I don't hit my daily token expenditure goal, so I have my own custom harness that actually hooks up an agent to TikTok, basically it splits up the reel into 1 second increments and then feeds those frames to the LLM for it's own consumption. I can easily burn 10m tokens a day on this, and Claude seems to enjoy it.

Personally I want to thank you Simon for putting me onto this "vibe engineering" concept, I really didn't expect an archaeology major like myself to become a real engineer but thanks to AI now I can be! Truly gatekeeping in tech is now dead.

For work I do agentic engineering. As the code that I submit for a code review is hand reviewed by me. I know every line and file that I submit.

My side project is 80% vibe code. Every now and then I look and see all the bad stuff, then I scold Codex a bit and it refactors it for me. So I do see the author's point.

Instead of "vibe coding" by asking the AI to design and write code, I'm having it refine my own designs, and write code under strict supervision and guidance, that I carefully review and iterate on.

I took a rock carving course in school that really enlightened me about software engineering, and it still applies today, especially to AI. You can't just decide what you want to carve, hold the chisel in just the right spot, and whack it with a hammer just perfectly so all the rock you want falls away leaving a perfect statue behind.

"I saw the angel in the marble and carved until I set him free." -Michelangelo

It's a long drawn out iterative process of making millions of tiny little chips, and letting the statue inside find its way out, in its natural form, instead of trying to impose a pre-determined form onto it.

Vibe coding is hoping your first whack of the hammer is going to make a good statue, then not even looking at the statue before shipping it!

But AI assisted conscientious coding (or agentic engineering as Simon calls it) is the opposite of that, where you chip away quickly and relentlessly, but you still have to carefully control where you chisel and what you carve away, and have an idea in your mind what you want before you start.

> I know full well that if you ask Claude Code to build a JSON API endpoint that runs a SQL query and outputs the results as JSON, it’s just going to do it right. It’s not going to mess that up. You have it add automated tests, you have it add documentation, you know it’s going to be good.

> But I’m not reviewing that code. And now I’ve got that feeling of guilt: if I haven’t reviewed the code, is it really responsible for me to use this in production?

Answer: it wholly depends upon what management has dictated be the goal for GenAI use at the time.

There seems to be a trend of people outside of engineering organizations thinking that the "iron triangle" of software (and really, all) engineering no longer holds. Fast, cheap, good: now we can pick all three, and there's no limit to the first one in particular. They don't see why you can't crank out 10x productivity. They've been financially incentivized to think that way, and really, they can't lose if they look at it from an "engineer headcount" standpoint. The outcomes are:

1) The GenAI-augmented engineer cranks out 10x productivity without any quality consequences down the line, and keeps them from having to pay other people

or

2) The GenAI-augmented engineer cranks out 10x productivity with quality consequences down the line, at which point the engineer has given another exhibit in the case as to why they should no longer be employed at that organization. Let the lawyers and market inertia deal with the big issues that exist beyond the 90-day fiscal reporting period.

Either way, they have a route to the destination of not paying engineers, and that's the end goal.

If you don't like that way of running a software engineering organization, well, you're not alone, but if nothing else, you could use GenAI to make working for yourself less risky.

Simon,

Just piggy backing on this post since I'm early:

Would love to see your take on how the AI and Django worlds will collide.

Honestly, I think the need for devs is total copium, the progress made in two years is astounding and in two years time they will be better at programming than 99% of programmers. It’s incredible what they can do now. No it’s not perfect but imagine where we’ll be in 5 or 10 years.

"Code quality" was always a mirage imo. Logic is what matters. I've used the internet from the early days, and probably 99% of software I used always had serious bugs. Ultima online was mentioned in HN recently: it was a real bug-and-exploit-fest. Banks, AAA games, companies like Uber with 1000's of engineers - they all had serious problems (and that's still true). It would be worst if some engineers didn't have that drive to code in high quality, but we gotta admit that was not ever enough. Even now with Claude Code, I see a lot of "specifications" that are far from specified enough - and people blame the LLM.

man i love this post

I'd be lying if I said I was not worried about the future. I am not necessarily worried in the sense that there will be some grave, impeding doom that awaits the future of humanity.

Rather, I just feel like I have to constantly remind myself of the impermanence of all things. Like snow, from water come to water gone.

Perhaps I put too much of my identity in being a programmer. Sure, LLMs cannot replace most us in their current state, but what about 5 years, 10 years, ..., 50 years from now? I just cannot help be feel a sense of nihilism and existential dread.

Some might argue that we will always be needed, but I am not certain I want to be needed in such a way. Of course, no one is taking hand-coding away from me. I can hand-code all I want on my own time, but occupationally that may be difficult in the future. I have rambled enough, but all and all, I do not think I want to participate in this society anymore, but I do not know how to escape it either.

I grew up on construction sites with my dad. If i've done well in my career, it was from watching him operate - managing huge construction crews, how he figured out who to put on what tasks, handling suprises, setbacks, all that stuff

My dad (now retired) was always super practical about stuff. He'd tell me pretty nonchalantly things like "yeah we're dealing with xyz constraint, we may have to cut a corner over here, but that's ok", when I asked him about it he gave me a little spiel that you can be thoughtful about how you do things, including when you can cut a corner and more importantly, what corners are ok to cut.

I really took that to heart - especially the "be thoughtful about the corners you cut"

If an LLM has consistently one shotted certain tasks and they are rote/mechanical - not reviewing that code is probably ok.

Are you getting lazy and not reviewing stuff that should be reviewed even if a human wrote it? That's probably not ok

I can live with some basic code that broke because it used outdated syntax somewhere (provided the code isn't part of a mission critical application), but I can't live with it fucking JWT signing etc

[flagged]

[dead]

[flagged]

[flagged]

[dead]

[flagged]

[dead]

[flagged]

[flagged]

[flagged]

[flagged]

[flagged]

[flagged]

[flagged]

[flagged]

[flagged]

[dead]

> And that feels about right to me. I can plumb my house if I watch enough YouTube videos on plumbing. I would rather hire a plumber.

I don't buy this argument at all. I think if we could pay $20/month to a service that would send over a junior plumber/carpenter/electrician with an encyclopedic knowledge of the craft, did the right thing the majority of the time, and we could observe and direct them, we'd all sign up for that in a heartbeat. Worst case, you have to hire an experienced, expensive person to fix the mess. Yes, I can hear everyone now, "worst case is they burn your house down." Sure, but as we're reminded _constantly_ when we read stories about AI agent catastrophes -- a human could wipe your prod database too. wHy ArE yOu HoLdInG iT tO a DiFfErEnT sTaNdArD???

The business side of the house is getting to live that scenario out right now as far as software goes. Sure you've got years of expertise that an LLM doesn't have _yet_. What makes you think it can't replace that part of your job as well?

Vibe coding is just coding now. Writing assembly used to be a thing too until higher and higher languages were created. LLM is like that except it compiles English to code. This scares lot of professionals understandably.

It is pure arrogance to expect that machines will never be able to code as good as a skilled human.

And AI generated code should be different than human code. AI has infinite memory for details. AI doesn’t need organizational patterns like classes. Potentially AI can write code that is more performant than any human.

Will it look like garbage? Sure. Will the code be more suited to the task? Yes.