This article resonated tremendously with me, speaking as someone who has been coding since the 1980s and had to develop my own taste the hard way, through lots and lots of mistakes. Whenever I see a software demo built by someone mostly using agents to do the work, I have to question whether it's any good on the inside - or whether any real intuition or judgment went into its design. On the flip side, maybe none of that matters anymore - as long as it works, who cares how it was built.
For majority of people, including business executives it's the latter. No one cares about artisan code. Software is just a means to an end for them and not an art form.
People treat their computers like any other appliance. They don't care if it's riddled with bugs, they'll gladly restart their computer or refresh the page at infinitum to get it to work.
I think it's also why marketing and designers get more of a say in the software than engineers. UI polish matters more since it increases sales.
This is the hardest thing to come into terms with in this industry.
I'm quite sad that I have to disagree with you and agree with GP: if quality siftware mattered to people, Microsoft would have gone out of business a long time ago.
If quality software mattered, javascript wouldn't be the most popular way to write desktop software.
If users cared at all about quality software, we wouldn't have 800MB chat apps.
Of course yes, and it's even easier in some cases. Since there's a language that defines the functionality (e.g: formal methods) that the agent can use/veriy against
Yes but that 10% is all that matter to the people that matter to the business. Like a car, most people care about how it looks and how they interact with it... the only people who care what is under the hood, for the most part, are other engineers.
I just want to add that programmers have only relatively recently become so abundant (i.e. replaceable) that they have so little say in how software should be built.
Programmers are still the experts in the room, but they don't get what they ask for for free anymore.
It resonated with me too, and even as a once fairly prolific coder that now uses agents to do all the coding, I feel like this is all relevant and does matter. The agent does whatever you ask and provides thin judgements based on what it thinks is your motivation and context, even feeding it the context it will often make bad generic choices. I think this is the part hard to optimize because it's not a verifiable reward, it's the feel of the software. I think Herdr is a great example, I've tried several dozen terminal multiplexers / session aggregators and they are all vibed and different, but this one felt so much better to use because of its choices and actually increased the number of parallel sessions I was able to keep going. My own projects I find I have to stress every detail, that is what makes the software good now is really diving into it, the experience feels like just as much of a craft as coding was and 'taste' captures the concept for me.
For what it is worth, I appreciate you sharing your experience. Looks like you have more programming experience than I have in life, so your comment matters to me. It also gets rather close to something I wanted to evaluate in the post: "whether any real intuition or judgment went into [its] design". I think that's a big question to dwell on these days.
Ironically, though, I had written a post that had people debating whether I put any intuition or judgement into. Would've been real ironic to instead add a post-scriptum that said "Yep, none of it matters. I just slopped something in 5 minutes and it easily went to the frontpage. Look how easy it is!" but unfortunately that is not why I write, not what I intended to achieve, and not something I am equipped to do. I care about my work, which is why I care about your comment as well.
“as long as it works” is doing a lot of work here. Did the people building it say “no, again” enough times, and at the right times, in order to make something that _actually_ works instead of _merely_ works?
People who care about their health aren't ignorant about that. I think that can apply to software too: if you care about safety or privacy you might think twice before using vibe-coded software.
That is a significant leap in logic. A vibecoding tool is not a McDouble. Nor is Chris Kempczinski the same person as Dario Amodei.
The biggest difference is that there are only so many ways to eat a McDouble, but you can actually wield a vibe coding tool with skill. There's nothing stopping you from looking at the code and asking for structural reworks or refactors that please your code taste. In fact one of the best use cases for AI coding tools is vibe refactoring old apps where the code is straight up horrible.
> but you can actually wield a vibe coding tool with skill. There's nothing stopping you from looking at the code
It’s hard to give something like “vibe coding” a rigid definition, but I think we can all agree that it’s inversely correlated with "looking at the code". The more I examine the code while building, the less the activity is about vibes.
So, used this way, it's true that it's not a McDouble anymore, but it's also not so much of a vibecoding tool anymore either.
i have never had a "mcdonald's mcdouble", whatever that is. i cook my own burgers because i am a human being and i want to eat real food, not nauseating slop that should only be fed to livestock
The LLM has its own intuition and judgement which at this point is quite good. I’m not just throwing that around, I use LLMs 12-14 hours a day and for a long time reviewed atleast 75% of the code.
I've spent a day fixing a feature my very junior colleague pushed, and which was merged by my should-have-known-better colleagues who have almost as much experience as me.
We have access to the same models. Why did her version suck and my features don't need fixing?
