Tip: I very often use AI for inspiration. In this case, I ended up keeping a lot (not all) of the UI code it made, but I will very often prompt an agent, throw away everything it did, and redo it myself (manually!). I find the "zero to one" stage of creation very difficult and time consuming and AI is excellent at being my muse.
This right here is the single biggest win for coding agents. I see and directionally agree with all the concerns people have about maintainability and sprawl in AI-mediated projects. I don't care, though, because the moment I can get a project up on its legs, to where I can interact with some substantial part of its functionality and refine it, I'm off to the races. It's getting to that golden moment that constitutes 80% of what's costly about programming for me.
This is the part where I simply don't understand the objections people have to coding agents. It seems so self-evidently valuable --- even if you do nothing else with an agent, even if you literally throw all the code away.
I was talking about this the other day with someone - broadly I agree with this, they're absolutely fantastic for getting a prototype so you can play with the interactions and just have something to poke at while testing an idea. There's two problems I've found with that, though - the first is that it's already a nightmare to convince management that something that looks and acts like the thing they want isn't actually ready for production, and the vibe coded code is even less ready for production than my previous prototyping efforts.
The second is that a hand-done prototype still teaches you something about the tech stack and the implementation - yes, the primary purpose is to get it running quickly so you can feel how it works, but there's usually some learning you get on the technical side, and often I've found my prototypes inform the underlying technical direction. With vibe coded prototypes, you don't get this - not only is the code basically unusable, but you really are back to starting from scratch if you decide to move forward - you've tested the idea, but you haven't really tested the tech or design.
I still think they're useful - I'm a big proponent of "prototype early," and we've been able to throw together some surprisingly large systems almost instantly with the LLMs - but I think you've gotta shift your understanding of the process. Non-LLM prototypes tend to be around step 4 or 5 of a hypothetical 10-step production process, LLM prototypes are closer to step 2. That's fine, but you need to set expectations around how much is left to do past the prototype, because it's more than it was before.
> the moment I can get a project up on its legs, to where I can interact with some substantial part of its functionality and refine it, I'm off to the races. [...] This is the part where I simply don't understand the objections people have to coding agents.
That's what's valuable to you. For me the zero to one part is the most rewarding and fun part, because that's when the possibilities are near endless, and you get to create something truly original and new. I feel I'd lose a lot of that if I let an AI model prime me into one direction.
It's not FUN building all the scaffolding and setting up build scripts and all the main functions and directory structures.
Nor do I want to use some kind of initialiser or skeleton project, they always overdo things in my opinion, adding too much and too little at the same time.
With AI I can have it whip up an MVP-level happy-paths-only skeleton project in minutes and then I can start iterating with the fun bits of the project.
> This is the part where I simply don't understand the objections people have to coding agents. It seems so self-evidently valuable --- even if you do nothing else with an agent, even if you literally throw all the code away.
It sounds like the blank page problem is a big issue for you, so tools that remove it are a big productivity boost.
Not everyone has the same problems, though. Software development is a very personal endeavor.
Just to be clear, I am not saying that people in category A or category B are better/worse programmers. Just that everyone’s workflow is different so everyone’s experience with tools is also different.
The key is to be empathetic and trust people when they say a tool does or doesn’t work for them. Both sides of the LLM argument tend to assume everyone is like them.
Just this week I watched an interview with Mitchell about his dev setup and when asked about using neovim instead of an IDE he said something along the lines of "I don't want something that writes code for me". I'm not pointing this out as a criticism, but rather that it's worth taking note that an accomplished developer like him sees value in LLMs that he didn't see in previous intellisense-like tooling.
Not sure exactly what you're referring to, but I'm guessing it may be this interview I did 2 years ago: https://youtu.be/rysgxl35EGc?t=214 (timestamp linked to LLM-relevant section) I'm genuinely curious because I don't quite remember saying the quote you're saying I did. I'm not denying it, but I'd love to know more of the context. :)
But, if it is the interview from 2 years ago, it revolved more around autocomplete and language servers. Agentic tooling was still nascent so a lot of what we were seeing back then was basically tab models and chat models.
As the popular quote goes, "When the Facts Change, I Change My Mind. What Do You Do, Sir?"
The facts and circumstances have changed considerably in recent years, and I have too!
They even used the quote as the title of the accompanying blog post.
As I say, I didn’t mean this as a gotcha or anything- I totally agree with the change and I have done similarly. I’ve always disabled autocomplete, tool tips, suggestions etc but now I am actively using Cursor daily.
Yeah understood, I'm not taking it negatively, I just genuinely wanted to understand where it came from.
Yeah this is from 2021 (!!!) and is directly related to LSPs. ChatGPT didn't even get launched until Nov 2022. So I think the quote doesn't really work in the context of today, it's literally from an era where I was looking at faster horses when cars were right around the corner and I had not a damn clue. Hah.
Off topic: I still dislike [most] LSPs and don't use them.
What do you not like about LSPs? When you do eg refactoring isn't it nice to do operations on something that actually reflects the structure of your code?
I use agents for that, and it does a shockingly good job. LSPs constantly take up resources, most are poorly written, and I have to worry about version compatibility, editor compatibility, etc. It's just a very annoying ecosystem to me.
External agent where I can say "rename this field from X to Y" or "rewrite this into a dedicated class and update all callers" and so on works way better. Obviously, you have to be careful reviewing it since its not working at the same level of guarantee an LSP is but its worth it.
The CEO of Sourcegraph Quinn was pretty negative on coding agents and agentic development only about 10 months ago [0]. He had 'agentic stuff' in the Deader category (Used rarely, Reviewing it aint worth it). In fairness, he did say it was the future but 'is not there yet'. Since then, Sourcregraph's code assistant plugin Cody has been deprecated an they are all in on agents and agentic with Amp.
Yeah, I said about coding agents, “it’s obviously the future, but it’s not there yet”. That talk was from the AI Engineer conference in June 2024 (16 months ago). Coding agents have come a long way since then!
Even then AI autocomplete is way better today than at the start of the year; never mind 2 years ago.
Not only are the suggestions better, but they are presented much less obtrusively now in IDEs. For example it no longer ambushes you with a 500 line suggestion while you are typing, it shows a bit of what it wants to add next and seems to pull away if you are laying down code while ignoring it.
> As the popular quote goes, "When the Facts Change, I Change My Mind. What Do You Do, Sir?"
> The facts and circumstances have changed considerably in recent years, and I have too!
This should be everyone's default if we are learning. Not saying that's what the OP did here, but I hate how people dig out something that was said years ago like it's some gotcha. I think it's a bigger issue if people never change their minds about anything because it would show hubris and a lack of learning.
I explain it to my peers as "exploiting Cunningham's Law[0] with thyself"
I'll stare blankly at a blank screen/file for hours seeking inspiration, but the moment I have something to criticise I am immediately productive and can focus.
About that quote, iirc it's also a technique in Intelligence for getting information from people. You say something stupid and wrong and they will instantly just correct you on the spot and explain why etc.
People get into this field for very different reasons.
- People who like the act and craftsmanship of coding itself. AI can encourage slop from other engineers and it trivializes the work. AI is a negative.
- People who like general engineering. AI is positive for reducing the amount of (mundane) code to write, but still requires significant high-level architectural guidance. It’s a tool.
- People who like product. AI can be useful for prototyping but won’t won’t be able to make a good product on its own. It’s a tool.
- People who just want to build a MVP. AI is honestly amazing at making something that at least works. It might be bad code but you are testing product fit. Koolaid mode.
That’s why everyone has a totally different viewpoint.
> - People who like the act and craftsmanship of coding itself. AI can encourage slop from other engineers and it trivializes the work. AI is a negative.
Those who value craftsmanships would value LLM, since they can pick up new languages or frameworks much faster. They can then master the newly acquired skills on their own if preferred, or they can use LLM to help along the way.
> People who like product. AI can be useful for prototyping but won’t won’t be able to make a good product on its own. It’s a tool.
Any serious product often comprises of multiple modules, layers, interfaces. LLM can help greatly with building some of those building blocks. Definitely a useful tool for product building.
> Those who value craftsmanships would value LLM, since they can pick up new languages or frameworks much faster. They can then master the newly acquired skills on their own if preferred, or they can use LLM to help along the way.
That's like saying those who value books would value movie adaptations because they can pick up new stories much faster.
Is it really so alien to you the someone might prefer learning a new language or framework by, gasp, reading its documentation?
Real subtle. Why not just write "there are good programmers and bad programmers and AI is good for bad programmers and only bad programmers"? Think about what you just said about Mitchell Hashimoto here.
I don't think it's an unfair statement that LLM-generated code typically is not very good - you can work with it and set up enough guard rails and guidance and whatnot that it can start to produce decent code, but out of the box, speed is definitely the selling point. They're basically junior interns.
If you consider an engineer's job to be writing code, sure, you could read OP's post as a shot, but I tend to switch between the personas they're listing pretty regularly in my job, and I think the read's about right.
To the OP's point, if the thing you like doing is actually crafting and writing the code, the LLMs have substantially less value - they're doing the thing you like doing and they're not putting the care into it you normally would. It's like giving a painter an inkjet printer - sure, it's faster, but that's not really the point here. Typically, when building the part of the system that's doing the heavy lifting, I'm writing that myself. That's where the dragons live, that's what's gotta be right, and it's usually not worth the effort to incorporate the LLMs.
If you're trying to build something that will provide long-term value to other people, the LLMs can reduce some of the boilerplate stuff (convert this spec into a struct, create matching endpoints for these other four objects, etc) - the "I build one, it builds the rest" model tends to actually work pretty well and can be a real force multiplier (alternatively, you can wind up in a state where the LLM has absolutely no idea what you're doing and its proposals are totally unhinged, or worse, where it's introducing bugs because it doesn't quite understand which objects are which).
If you've got your product manager hat on, being able to quickly prototype designs and interactions can make a huge, huge difference in what kind of feedback you get from your users - "hey try this out and let me know what you think" as opposed to "would you use this imaginary thing if I built it?" The point is to poke at the toy, not build something durable.
Same with the MVP/technical prototyping - usually the question you're trying to answer is "would this work at all", and letting the LLM crap out the shittiest version of the thing that could possibly work is often sufficient to find out.
The thing is, I think these are all things good engineers _do_. We're not always painting the Sistine Chapel, we also have to build the rest of the building, run the plumbing, design the thing, and try to get buy-in from the relevant parties. LLMs are a tool like any other - they're not the one you pull out when you're painting Adam, but an awful lot of our work doesn't need to be done to that standard.
First, I agree with tptacek that Ghostty is a work of craftsmanship. Mitchell is a very talented dev and he says he greatly benefits from using AI.
On the other hand I understand your point that some people got into coding because of coding and they like doing that manually. Unfortunately, we're not being paid to do what we like, but to solve problems with code. What we like is usually a hobby. Software engineering had a golden run for 20-30 years where we were paid well to do things we enjoyed doing, but unfortunately that might change. As an analogy, think about woodworking: there's craftsmanship in a nice wood table, but at the end of day I won't pay thousands of dollars for one, when a couple of hundred dollars will buy me a good enough one from IKEA (maybe you're not like that, but the general population is).
I love coding, and I'm extremely picky, but I also agree with you it's not what I'm being paid to do. And even in hobby projects there are plenty of rough drafts necessary that I can outsource to an LLM, and then I get to refine things afterward, when they're working, and I can obsess over details...
To take your table analogy, which I fully agree with, if I wanted to take the effort of crafting a nice wooden table, I'd happily have someone cut the pieces for me and do the basic stuff that doesn't require specific skills, and then spend my time applying the finer details that makes the difference between a basic table and a great table. I get that some people would want to do everything "from scratch", but I'd rather focus on where I can make the most difference.
I can't get past the framing that "people who like the act and craftsmanship" feel AI is negative, which implicitly defines whatever Mitchell Hashimoto is doing as not craftsmanship, which: ghostty is pure craftsmanship (the only reason anyone would spend months writing a new terminal).
No, I think my response was fair, if worded sharply. I stand by it.
Your response conflates the categories of people and ignores that statements like "people..." can mean "(some|most|all) people..." in casual writing/speech. It is not a fair response.
I agree with your frustration with the framing that _all_ people who like the act and craftsmanship feel AI is negative, and the consequence that if one does like AI, then they must not like the act of craftsmanship. Many such people view it as a tool, Mitchell included.
No, it didn't mean that. It mean what I thought it meant. It was a roundabout way of saying that people who use AI aren't interested in craft. The meaning was plain.
They're far more than junior interns. I've had a long languishing project to build an ahead-of-time Ruby compiler. It started as a toy, and a blog series, and I then mostly put it on ice about a decade ago, except for very occasional little rounds of hacking. It self-hosts, but is very limited.
A week or so ago, I gave Claude a task of making it compile rubyspecs. I then asked it to keep making specs pass. I do need to babysit it, but it's doing debugging and work no junior I've ever worked with could be trusted to do. It knows how to work with gdb, and trace x86 assembler. It understands how to read the code of a complex compiler, and modify code generation and parsers that even I - who wrote it in the first place - sometimes find challenging.
It's currently (as I'm writing this) working its way through adding bignum support. Which in Ruby is tricky because it now no longer splits it in two classes - the code need to handle tagged integers that gets auto-promoted to heap allocated objects, that to the user has the same class. I spent the morning swearing at it, but then reset with a clearer description and it produced an extensive plan, and started working through it.
I'll agree it's not great code without a lot of coaxing, but it's doing stuff that even a lot of senior, highly experienced developers would struggle with.
I will agree it needs oversight, and someone experienced guiding it, like a junior developer would, but if I had junior developers producing things this complex, I'd lock them in a basement and never let them go (okay, maybe not).
One of the hardest things, I find, where I will agree it smells of junior developer sometimes, is that it's impatient (once it even said "this is getting tedious") and skipping ahead instead of carefully testing assumptions and building up a solution step by step if you don't tell it very clearly how to work.
I don't think we disagree that much, btw., I just wanted to describe my recent experience with it - it's been amazing to see, and is changing how I work with LLMs, in terms of giving it plenty of scratchpads and focusing on guiding how it works, making it create an ambitious plan to work to, and getting out of its way more, instead of continuously giving it small tasks and obsessing over reviewing intermediate work product.
What I'm seeing often with this approach is that whenever I see something that annoys me scroll past, it's often fixed before I've even had a chance to tell it off for doing something stupid.
> This is the part where I simply don't understand the objections people have to coding agents
Because I have a coworker who is pushing slop at unsustainable levels, and proclaiming to management how much more productive he is. It’s now even more of a risk to my career to speak up about how awful his PRs are to review (and I’m not the only one on the team who wishes to speak up).
The internet is rife with people who claim to be living in the future where they are now a 10x dev. Making these claims costs almost nothing, but it is negatively effecting mine and many others day to day.
I’m not necessarily blaming these internet voices (I don’t blame a bear for killing a hiker), but the damage they’re doing is still real.
I don't think you read the sentence you're responding to carefully enough. The antecedent of "this" isn't "coding agents" generally: it's "the value of an agent getting you past the blank page stage to a point where the substantive core of your feature functions well enough to start iterating on". If you want to respond to the argument I made there, you have to respond to the actual argument, not a broader one that's easier (and much less interesting) to take swipes at.
I have to agree. My experience working on a team with mixed levels of seniority and coding experience is that everybody got some increase in productivity and some increase in quality.
The ones who spend more time developing their agentic coding as a skillset have gotten much better results.
In our team people are also more willing to respond to feedback because nitpicks and requests to restructure/rearchitect are evaluated on merit instead of how time-consuming or boring they would have been to take on.
> My experience working on a team with mixed levels of seniority and coding experience is that everybody got some increase in productivity and some increase in quality.
Is that true? There have been a couple of papers that show that people have the perception that they are more productive because the AI feels like motion (you're "stuck" less often) when in reality it has been a net negative.
Don't mention AI, just point out why the code is bad. I've had co-workers who were vim wizards and others who literally hunt and pecked to type. At no point did their tools ever come up when reviewing their code. AI is a tool like anything else, treat it that way. This also means that the OPs default can't be AI == bad; focus on the result.
I don't think it is a long term solution. More like training wheels. Ideally the engineers learn to use AI to produce better code the first time. You just have a quality gate.
Edit: Do I advocate for this? 1000%. This isn't crypto burning electricity to make a ledger. This objectively will make the life of the craftsmanship focused engineer easier. Sloppy execution oriented engineers are not a new phenomenon, just magnified with the fire hose that an agentic AI can be.
The environmental cost of AI is mostly in training afaik. The inference energy cost is similar to the google searches and reddit etc loads you might do during handwritten dev last I checked. This might be completely wrong though
I hear this argument a lot, but it doesn’t hold water for me. Obviously the use of the AI is the thing that makes it worthwhile to do the training, so you obviously need to amortize the training cost over the inference. I don’t know whether or not doing so makes the environmental cost substantially higher, though.
> If you can describe why it is slop, an AI can probably identify the underlying issues automatically
I would argue against this. Most of the time the things we find in review are due to extra considerations, often business, architectural etc, things which the AI doesn't have context of and it is quite bothersome to provide this context.
I generally agree that vague 1 shot prompting might vary.
I also feel all of those things can be explained over time into a compendium that is input. For example, every time it is right, or wrong, comment and add it to an .md file. Better yet, have the CLI Ai tool append it.
We know what is included as part of a prompt (like the above) is more accurately paid attention to.
My intent isn't to make more work, it's just to make it easier to highlight the issues with code that's mindlessly generated, or is overly convoluted when a simple approach will do.
I'm the opposite, I find getting started easy and rewarding, I don't generally get blocked there. Where I get blocked, after almost thirty years of development, is writing the code.
I really like building things, but they're all basically putting the same code together in slightly different ways, so the part I find rewarding isn't the coding, it's the seeing everything come together in the end. That's why I really like LLMs, they let me do all the fun parts without any of the boring parts I've done a thousand times before.
It's funny because the part you find challenging is exactly the thing LLM skeptics tend to say accounts for almost none of the work (writing the code). I personally find that once my project is up on its legs and I'm in flow state, writing code is easy and pleasant, but one thing clear from this thread is everyone experiences programming a little differently.
Yeah, definitely. I do agree with the skeptics to a point, as I don't let the LLM write code without reviewing (it makes many mistakes that compound), but I'd still rather have it write a function, review and steer, have it write another, and so on, than write database models myself for the millionth time.
It's not that I find it hard, I've just done it so many times that it's boring. Maybe I should be solving different/harder problems, but I don't mind having the LLM write the code, and I'm doing what I like and I'm more productive than ever, so eh!
I was just talking about this in a chat today, because 'simonw had at some point talked about getting to the point where he was letting go of reviewing every line of LLM code, and I am nowhere close to that point --- I'll take Claude's word on Tailwind classes as long as the HTML looks right, but actual code, I review line-by-line, token-by-token, and usually rewrite things, even if just for cosmetic reasons.
I've had Claude write ~50+ commits of compiler code over the last week, and I've mostly not even read it yet. I will review the end result at some point, and I do spot checks, but for the most part I review whether or not the number of specs that passes is going the right way, and occasionally butt in when I glance over and spot it doing something silly.
I fully get why people won't trust it, but I'm leaning more and more on reigning it in with specs and tests and process, and then letting it run for extended amounts of time, and then giving it instructions to fix obvious issues, before finally reviewing the result, and letting it fix whatever comes up.
Yeah, there's a definite continuum for where LLMs are most to least expert. They seem to be fairly OK with JS, less so with Python, and C is just a crapshoot.
It depends on the project as well, for throwaway things I'm fine to just let it do whatever it wants, but for projects that I need to last more than a few days, I review everything.
A friend of mine has been doing embedded C stuff (making some kind of LCD wall) and has been blown away by how well it's been doing with C --- he went in an LLM skeptic (I've been trying to sell him for months, it finally clicked).
Huh, interesting. I wanted to turn an old rotary phone into a meeting headset, and I tried to get Opus to make me a sound card, but $35 in API costs later, I had no sound card.
> the moment I can get a project up on its legs, to where I can interact with some substantial part of its functionality and refine it, I'm off to the races
AI is an absolute boon for "getting off the ground" by offloading a lot of the boilerplate and scaffolding that one tends to lose enthusiasm for after having to do it for the 99th time.
> AI is excellent at being my muse.
I'm guessing we have a different definition for muse. Though I admit I'm speaking more about writing (than coding) here but for myself, a muse is the veritable fount of creation - the source of ideas.
Feel free to crank the "temperature" on your LLM until the literal and figurative oceans boil off into space, at the end of the day you're still getting the ultimate statistical distillation.
Agree – my personal rule is that I throw away any branches where I use LLM-generated code, and I still find it very helpful because of the speed of prototyping various ideas.
I concur here and would like to add that I worry less about sprawl when I know I can ask the agent to rework things in future. Yes, it will at times implement the same thing twice in two different files. Later, I’ll ask it to abstract that away to a library. This is frankly how a lot of human coding effort goes too.
This is an artefact of a language ecosystem that does not prioritize getting started. If you picked php/laravel with a few commands you are ahead of the days of work piping golang or node requires to get to a starting point.
I guess it depends on your business. I rarely start new professional projects, but I maintain them for 5+ years - a few pieces of production software I started are now in the double digits. Ghostty definitely aims to be in that camp of software.
This right here is the single biggest win for coding agents. I see and directionally agree with all the concerns people have about maintainability and sprawl in AI-mediated projects. I don't care, though, because the moment I can get a project up on its legs, to where I can interact with some substantial part of its functionality and refine it, I'm off to the races. It's getting to that golden moment that constitutes 80% of what's costly about programming for me.
This is the part where I simply don't understand the objections people have to coding agents. It seems so self-evidently valuable --- even if you do nothing else with an agent, even if you literally throw all the code away.
PS
Put a weight on that bacon!