[goopthink]

Why couldn’t OpenAI vibe code their own Windsurf/Cursor competitor? (Serious question).

OpenAI is a technology company constantly in search of productization (ChatGPT, Sora, Dall-e), and they’ve been really good at creating product interest that converts to acquisition. An IDE is much more complex than a chat app, but given their literal billions of dollars and familiarity with developer tooling, this is a down-stack build that they could dogfood off their own tech. And especially given that some of these tools were built by tiny teams (Cursor is what, 10 people?), is this like Google and Facebook’s implicit admission that they can’t “build and grow” anymore, and need to turn to acquisitions to fuel growth?

[conartist6]

I don't think there's anyone better positioned to answer your question than I am, given that I've spent the last 3 years building the IDE tech that OpenAI really wants and needs right now (though they don't yet know it).

The problem is that what you're discussing is a political undertaking. I don't mean that it's "left versus right" political, I mean that the primary task that makes it hard is getting a lot of human beings to agree on some low-level details about how the gory internals of an IDE work. LLMs can produce text, but they have no will to political organization. They aren't going to accomplish a task by going out and trying align the needs of many individuals in a compromise that requires determined work to find out what those people really want and need, which is the only way to get this particular task done. Somehow the natural-language-as-API idea has made many people think that in the future APIs and formal technical standards will be unnecessary, something for which I do not see evidence.

There's a second problem too, one of alignment: LLMs encourage you to give away the work of coding. I would not for anything give away the pain of using mediocre tools like VSCode to write a lot of code myself, because what I learn as the intelligence doing the work is what I need to know to be able to make tools for doing that work more efficiently. I use my learnings to design protocols, data structures, libraries, frameworks, and programming languages which don't just *mask* the underlying pain of software development but which can reduce it greatly.

[surgical_fire]

> Why couldn’t OpenAI vibe code their own Windsurf/Cursor competitor?

The obvious answer is that vibecoding does not work.

If it did, OpenAI wouldn't need to buy Windsurf

[conartist6]

I'm as much of a skeptic about vibe coding as you are. I consider engineering a bit of an art, so it's not for me.

That said, the kinds of things I've been hearing as going wrong with vibe coding are all or mostly fixable. I would even go as far as to say they're manifestations of the same problems that make IDEs such clumsy tools for humans to use at the moment, which is why we're trying to foist all this annoyance of using them off onto AIs.

The biggest problem, THE problem in this space is that tools we use don't have the language to describe the logic -- the intent -- behind refactorings. If you've never captured the user's intent, you can't learn from it. The ability to capture that intent is absent all the way from the highest level UX right down the lowest level of serialization: the patch format

That's why I believe it's my tech they really need and not Windsurf's -- I'm working on capturing the intent top to bottom.

[skydhash]

Isn’t the intent described in the commit message, which is one of many ways the developer can use to share intent? Code is only the what and the how. The why should be in Wiki, Readmes, Commit, Issue tracker…

[conartist6]

Let's say the change is to rename a variable. The patch format can never capture that the intent was to rename a variable. You can say so in a commit message, sure, but the action is never recorded, only it's effects, causing merges involving the rename to have conflicts

[skydhash]

That’s because the code is data for an execution machine. All the other stuff are for the hunan mind. There’s multiple way to transition from one state of code to another, so mostly people record the existence of a transition (version control) and why it has happened (commits and other texts).

Recording how is not fruitful, unless the technique is good. In this case, the essence will be extracted, and it will become a checklist or a script.

If you have two itents producing conflicting patches, the merge intent that emerges is good because it brings empathy (understanding the message) and harmony (working code). And that’s why almost everyone says that the code itself doesn’t matter. Only the feature that it brings into play and the the mind that birth it do. It is a medium.

And a merge conflict is nice because it detects overlapping work and the fact that the concerned people should propably meet.

[mycall]

Typically intent needs to be well defined, such as PR discussions, use case / need assessments, or user stories on a backlog.

[hotstickyballs]

Intent being unclear is exactly why different languages exist

[re-thc]

> the kinds of things I've been hearing as going wrong

That’s the problem with this stuff. It’s all hearsay.

[cdelsolar]

Iunno, Cline is vibe coding some stuff for me right now while I’m surfing through hacker News.

[conartist6]

Yeah that's weird to me but I've made my peace with it. I do this to make the software I want to use, and I think that's a chance everyone should have.

[pomatic]

This is patently wrong - as an experiment, I've vibe coded* four apps in the last 10 days. I did not write one line of code myself. Two are crud style, one is DNS related and one is for an embedded device. They were non-trivial use cases, built using claude desktop with various MCP tools. I was blown away by just how good it was.

*the original definition of vibe coding involved using voice to dictate the prompts to AI, I prefer to type.

[gdilla]

I mean, why did Facebook buy Instagram? It's a feed of photos and social network, hmmmm. In one sense, sure you can copy any software out there when you have the resources, but OpenAI got to peak under the hood of Windsurf and 1) Found a large, growing community (emphasis on growing) they can buy, which is worth something and 2) A team that probably will save them 1 year of unforced errors (also worth something) and 3) they have a lot of capital to deploy anyway, and this is not a risky bet.

[sazoo]

Exactly. Some of the ultra-myopic takes regarding vibe coding reek of "the lady doth protest too much" lol. Cope on!

[sazoo]

Your comment hurts my brain. Help me understand your reasoning here:

"Technology no worky. Therefore let's spend $3billion to buy one of the leading tools that already has a million users."

Did you even stop for 5 seconds to think about how completely nonsensical your comment is? The dev-cope is strong here... XD

[surgical_fire]

I am not the one coping. Many things may threaten my earnings as a Dev, mostly the very clear possibility of a difficult economic scenario in the years to come.

Vibe coding is just people deeply invested in a deadend technology trying to hype it up.

And I say this as someone that does use AI to a limited capacity as a code assistant.

[conartist6]

And the next link I click captures it perfectly: https://blog.ollien.com/posts/llm-friction/

To quote Ollien:

> Even as an “LLM-skeptic”, it would be silly for me to say that the tools are useless for software development. There are clearly times where they can be useful, whether that’s to perform a refactoring too complicated for IDE tooling, or to get a proof-of-concept put together. With that said, in my time both using and watching others use LLMs, I have noticed a troubling trend: they help reduce friction when developing software – friction that can help us to better understand and improve the systems we work on.

It's quite funny to me that in the case I'm talking about, "a refactor that's too complicated for IDE tooling" is exactly the kind of friction that needs to be felt

[dickersnoodle]

If the refactoring is beyond the IDE tooling it's even more incumbent on the engineers/developers to know what they're doing instead of outsourcing it to the "brains" behind an LLM.

[never_inline]

I wouldn't use LLM for serious refactors. It's actually the common "recipes" using popular libraries where LLM is most useful for me now.

[__loam]

That friction is the smell you need to chase down for a refactor.

[jv22222]

I’ve been building text editor as notion from the ground up. Recently i’ve been uploading the full codebase to Gemini to discuss different tasks. I can fully agree that the LLM can simply not decide on how to do things! If you slightly tweak your prompt it will suggest an entirely different architecture. You have to be very careful about pulling out the right stuff from the cargo cult best practice stuff baked in to the model. Another thing, a really obvious optimization might be just staring it in the face and it won’t suggest it unless you say “have you considered XYZ?” and then all of a sudden it won’t stop talking about XYZ! But most of all you just have to be strong in your will power to guide it in the direction you want.

[runfaster2000]

This excellent presentation (Veritasium) is making much the same point (learning comes from “reps”).

https://youtu.be/0xS68sl2D70?si=rXg5_wjg4bEdMalw

[lozenge]

Why would OpenAI vibe code anything? They have hundreds of experienced software engineers. Vibe coding is meant to replace an intern, not your actual team.

They don't want to get into specific industries (AI for software development, AI for business process management, AI for knowledge workers). They just do the AI component. Maybe that's changing now, but they still let other people take the risks - you know, people who are passionate about coding. Then when the product is proven, they might acquire it.

[echoangle]

Why would you replace an intern with AI? The goal is to give someone experience and make them come back for a job later. If you replace interns with AI, don’t complain about not getting developers later.

[MonkeyClub]

Exactly: without interns and juniors now you won't have mediors and seniors later.

[lumost]

My honest take? Vibe coding makes a senior engineer 2-4x more productive depending on the project. Very large projects see diminishing returns down to 0% productivity gain or negative. I can probably supervise 10 concurrent vibe coding sessions with a little thought on how to structure the tasks and code. This is like giving each of your top engineers their own staff.

The AI coders are different from human coders in what they can and can’t do, they are both profoundly dumb - and extremely technically proficient *at the same time*.

[__loam]

Insane take. 2-4x productive until you have to refactor or fix anything and realize you just created a 10,000 line steaming pile.

[mitemte]

I recently spent 2 weeks fixing a project that a senior engineer seemingly vibe coded while I was on holiday. Prior to that, their work output was excellent in terms of quality and pace.

Those 2 weeks were absolute hell for me. I estimate I had to rewrite about 90% of the code. Everything was cobbled together and ultimately disposable. Unfortunately, this work was meant to be the first of several milestones and was completely unsalvageable as a foundation for the project.

I'm not opposed to using AI tools, I use them myself. But being on the receiving end and having to deal with someone else's vibe coded rubbish is truly dreadful.

[__loam]

I am opposed to them. I'm tired of being pushed by people who don't understand the profession to use crappy tools to be unrealistically productive. I hate what their presence has done to the industry and to the expectations placed on us.

[monkeycantype]

I disagree with this negative take. I can use Claude to quickly explore libraries, I’m not familiar with, and have developed a development process where I describe the purpose of each class and method in a markdown file , and have Claude, Gemini, deep seek and Chat all pitch descriptions of how to implement it in shared markdown files. I correct their misconceptions and inefficiencies before any code is written, I can write this code myself, but I’m finding I can work faster like this.

[conartist6]

So you've managed to ensure that the one things you never have to do is come up with any ideas. You're turning yourself into the robot

[misfits]

That's cool, I will try it!

[__loam]

So when Claude lies to you what do you do? Your workflow is crazy. Just write the fucking code.

[monkeycantype]

Importantly this is how I work on my own personal fun projects, so it really doesn't matter if is productive, I find it enjoyable so I'm going to keep exploring it, there will be pros and cons, don't have the final ficture on that yet

claude can't manage the big picture of what I'm trying to achieve, and claude and the others hallucinate all the time.

I have them all write simple tests for the code, so if they introduce me to a new library, I have tests to prove their assumptions.

And I review everything and tweak everything.

In my day job I work as a lead dev / arch, this isn't must different, working with Claude is like working with a large team of very inconsistent devs, with deep knowledge, but a tenous grasp on reality that struggle with attention. So not that much different from real people?

My dad wrote code generators, back in the 70's and 80's that I did some work on early in my career, those code generators which took a high level description of a program and output mainframe code, made most of the money that paid for raising my siblings and I. From that perspective I've been roboting myself my entire professional life.

[whatevaa]

Found the manager who doesn't code.

[ramesh31]

They couldn’t care less about the tech. This was entirely about corporate accounts and user data. Windsurf caught the first wave of enterprise adoption to agentic coding IDEs, and they have tons of big customers now.

[goopthink]

No disagreement, I think developer augmentation is an amazing productization of LLMs, and will likely be better at converting enterprises so paying customers.

But OpenAI has the best brand recognition and the largest user base, and they have the core tech powering all of this. Whats the number on “tons” of customers, given that these VSCode-spinoff/plugin GPT wrappers are sprouting around like mushrooms after a rain?

If this is a build-vs-buy decision, $3 billion? Is that worth 1/3rd of the money in the bank when they’re burning cash at insane rates just running servers, and the rest of the $30b fundraise is tenuous and there may not be a followup? I’m skeptical of the financial decision here.

[ramesh31]

OpenAI has massive B2C brand value, but their enterprise offerings are scant. The era of consumer chatbot AI is coming to a close, and the winner will be the one who captures the B2B mindshare for real applications.

It’s about buying time as well. Could they put out a competitive product in less than six months? Maybe, but even that would leave them light years behind the market by the time it was ready.

[Terretta]

Interesting since OpenAI and Anthropic both still mostly refuse to talk to anyone with "corporate" (information security, AAA, audit log) needs for under 150 seats even if they're willing to pay.

[mycall]

Microsoft, Oracle, Goole and the like are happy to handle corporate seats.

[siva7]

Vibe coding with what? o4-mini? 4.1? Their coding models are a joke and their agent coder product as well. They would have to use claude

[jononor]

There seems to be no stable consensus on which LLM one should have used, to get good results. Which is somewhat natural, things are moving quickly - and evaluation methods are immature (and the little we have, actively gamed).

But a lot of the arguments seem on the surface to be of "No true Scotchman AI" form. Or "you are just holding it wrong" (ref Apple).

[cdelsolar]

I find 4.1 superior to Claude in my few adventures into vibe coding.

[siva7]

Which tool do you use?

[cdelsolar]

Cline

[LZ_Khan]

Well I assume that's what Anthropic and Google are doing with Claude Code and Firebase Studio.

The main thing is marketing and userbase, which shouldn't be underestimated.

[ZeroTalent]

Cash is cheap. $3B isn't that much in the grand scheme of things for OpenAI and, allegedly, 1M users.

[rvz]

> Why couldn’t OpenAI vibe code their own Windsurf/Cursor competitor? (Serious question).

Because an IDE has to be tested to work and function correctly. Not "Vibe coded".

Vibe-coding is not software engineering.

It is better to build than to buy.

[ivanbalepin]

it still takes time to spec out, build and sell to a comparable size user base, even with 10 people. And you're not guaranteed the same results if you just try to clone all of that. If the price is right, why not take the shortcut.

[goopthink]

Is $3 billion the right price? That might be what Windsurf is being valued at (cue the “selling to willing buyers at currently fair market prices” meme), but that’s like saying “it would cost OpenAI more than $3b to staff from zero, build a competitor, and acquire a comparable volume of paying users” … and that feels like an insane statement given the implications therein?

Especially given that Windsurf (and I think Cursor too) is a VSCode fork and OpenAI is cozy enough with Microsoft? It’s not even a zero to one build.

[ivanbalepin]

if they are optimizing for cost, which they are obviously not, then of course it would take them less to build. If they are optimizing for time, and the actual aforementioned "vibe code" step may not even be the most time-consuming part, then yes, it may be the right price.

[echan00]

They could. They prob need talent though, considering how much fish they're frying

[1zael]

They are just too focused on more important problems to solve, primarily around model improvement and getting to AGI.

[fullstackwife]

Anyone can vibe code their own Windsurf/Cursor alternative, and adjust it accordingly to their needs. You don't need majority of the features, most of them are developed because of the usual "large enterprise customer wants a feature" mantra. The real value comes from the underlying model anyway.

This looks more like a short term tactical move, and the goal is to improve their "API usage" KPIs, because compared to Gemini/Claude it looks not good. (majority of traffic are chatgpt.com free users)