> pointing out differences between the thing I worked on for 24 hours and a room designed painstakingly by hand over weeks.
I went in expecting something handmade, and instead got something that was Claude-made. This doesn’t make me feel good, or feel user-centric. Now we’re throwing Blender into a webpage and serving it client-side rendered to users?
If anything, all of this feels more like bloat rather than something personal. It’s art for Claude’s sake. It’s not telling a story. Even on the actual blog post, the dropdown table of contents displays how long it takes to read each section. It feels so disconnected because it’s possible to throw a feature in, but is it really necessary? Maybe I am getting better at determining AI-made UIs, but all of this feels incredibly artificial.
Lastly, for all of the parallel QA checks and threads, all of the header links besides the “About” page throw a 404 at me. It’s almost ironic.
Yes, I agree. It only having one working link is the killer to me as well, because no one checked it besides the pretty front page. There was an article here yesterday about recognizable signs of AI-designed UIs and it was accurate in terms of what most of these sites now look like.
Ironically, this is the only part I was supposed to do by hand and didn’t give to Claude - making sure the HTTP redirects are setup correctly. Can’t believe I didn’t notice that. Anyway, you can blame dumb dumb me, not Claude.
I think we probably disagree on what level of abstraction means “personal”, but this website feels deeply personal to me. It existed in my head before it existed on a computer. I understand the argument that any form of expression needs to involve labor closer to the raw source material, but then again, I’ve never bought the argument that painters need to mix their own colors for a painting to be theirs. Having worked various layers of the web rendering stack, it also never made a lot of sense to me that you must use a certain layer to be considered “low enough”. Whenever someone claims that it’s JavaScript in Three.js, I immediately suspect, maybe unfairly, that the person has only the faintest idea of how many systems turn their loose instructions and turned them into pixels on their screen.
I don’t doubt that there is technical feat and prowess in being able to do it, and if this satisfies your vision, then you were able to contract an AI agent to do in full, but it is still a layer away from the personal effects of art done by one’s self. I just feel that this still leans more towards some new type of commission.
But largely more is this is a display of how we as people view what AI _seems_ capable of, and how we aren’t being fully aware of what actually has been accomplished. I appreciate the willingness to put this out for the world to see, criticism or not, yet if this is a reflection of how you want your website to be perceived, then having it in a ready, working state tells me how carefully you coordinated the effort. Right now, it just feels reinforces the narrative that work with AI has lost a large sense of quality, even if in this case, it didn’t handle that part. I suppose I expected the article to be something different, but that is on me.
> One of the foremost tasks of art has always been the creation of a demand
which could be fully satisfied only later. The history of every art form shows critical epochs in which a certain art form aspires to effects which could be fully obtained only with a changed technical standard, that is to say, in a new art form. The extravagances and crudities of art which thus appear, particularly in the so-called decadent epochs, actually arise from the nucleus of its richest historical energies. In recent years, such barbarisms were abundant in Dadaism. It is only now that its impulse becomes discernible: Dadaism attempted to create by pictorial—and literary—means the effects which the public today seeks in the film.
> Every fundamentally new, pioneering creation of demands will carry
beyond its goal.
Edit: to put a finer point on it, AI art like this is, through it's rigorous application in almost every previous medium, showing that there is a demand and energy for a new art form that depends on the changing technical standard and capabilities that AI has introduced. I don't know what it will be, but you can see some of the edges of it with how it is making waves in all the other mediums.
Interesting write up, the trick to diff renders to have agents check for regression during optimization is clever. The whole thing VERY much scans as an ad to me though. I like company tech demos but I'm not huge on reading to the end of a post talking about how awesome anthropic's walled garden is to read that he's an employee.
Also, the actual website fully crashed Firefox on my phone after making it play horrible grinding sounds over my headphones for like 30 seconds. That's a new one, even for heavy vibe code.
Great article, though if the author ever reads this: please choose "artificial", next time! It's already the first half of AI. If I'm bound to be an irrelevant spec in an irrelevant backwater of an irrelevant galaxy in the moments between abiogenesis and abioterminus, I can at least hope to make one obvious adjective a "thing".
That one Anthropic engineer made "vibecoding" happen overnight, so I write this with full confidence that they have magical influence over the ever-forming neologisms.
One small correction: Andrej Karpathy coined the term "vibe coding" in February 2025, a year after leaving OpenAI for the second time. He joined Anthropic in May 2026.
> I started by creating a new Claude Project. Projects are becoming a layer that spans Chat, Cowork, and Code; you can think of them as sitting above domain-specific concepts. You talk to one Claude with the ability to spin off additional sessions, share files and memory between them, and supervise the entire fleet of Claudes to achieve goals.
> Earlier this year, I would have joked that I get a lot done by telling a bunch of Claudes to do work. Now, I’ve put one Claude in charge of each of my projects, and in turn, that Claude knows how to manage tasks, investigations, and threads. We’re talking more about goals and less about how to achieve them.
My little Project setup.
> Every thread gets its own computer in the cloud, where it installs whatever it needs (Blender, FFmpeg, Playwright), checks its work with screenshots from a headless browser, and publishes a private preview page for me to click through. > All threads share one folder of files and one memory. Claude is smart enough to write down decisions I made in memory and makes sure newly started threads follow them without me having to repeat myself.
I must be missing the setting in my claude/claude code - how does one do that?
A project is just a custom prompt and some files you can attach.
Claude code is the same thing with some .claude files, and the folder directory is local to your machine.
Where is the ON switch for all of this? Showing the method and steps and prompts would be more useful than this post.
It appears that they're replacing the "old" version of projects you describe, with a new version mentioned in the post. It's being rolled out in stages:
It's all getting abstracted away.Even our computers are.But it's cool to see people continue to pushing the envelope.With the latest models releases, I'm starting to think about what even are humans for now.I guess we will continue evolving and keeping track but with how the models are competent now, I can already be satisfied for the foreseeable future.I don't think I would need anything more.And yet we will continue to push.
Super cool, love the aesthetic. Buy testing the whole blog a bit better would've surfaced the "all posts" link going to an empty page I guess? Also: not reading code for a toy project like this vs in a commercial setting shouldn't be compared at all.
Would also really love to see the token spend on this.
Yes! That is what I came to say: it's interesting, but I can't contextualize the workflow he describes without knowing the the quantity of tokens involved and their retail cost.
> pointing out differences between the thing I worked on for 24 hours and a room designed painstakingly by hand over weeks.
I went in expecting something handmade, and instead got something that was Claude-made. This doesn’t make me feel good, or feel user-centric. Now we’re throwing Blender into a webpage and serving it client-side rendered to users?
If anything, all of this feels more like bloat rather than something personal. It’s art for Claude’s sake. It’s not telling a story. Even on the actual blog post, the dropdown table of contents displays how long it takes to read each section. It feels so disconnected because it’s possible to throw a feature in, but is it really necessary? Maybe I am getting better at determining AI-made UIs, but all of this feels incredibly artificial.
Lastly, for all of the parallel QA checks and threads, all of the header links besides the “About” page throw a 404 at me. It’s almost ironic.
The website, to me, looks like Claude's project with sprinklings of the author's intent, and not the other way around.
Yes, I agree. It only having one working link is the killer to me as well, because no one checked it besides the pretty front page. There was an article here yesterday about recognizable signs of AI-designed UIs and it was accurate in terms of what most of these sites now look like.
https://news.ycombinator.com/item?id=49867038
Ironically, this is the only part I was supposed to do by hand and didn’t give to Claude - making sure the HTTP redirects are setup correctly. Can’t believe I didn’t notice that. Anyway, you can blame dumb dumb me, not Claude.
I think we probably disagree on what level of abstraction means “personal”, but this website feels deeply personal to me. It existed in my head before it existed on a computer. I understand the argument that any form of expression needs to involve labor closer to the raw source material, but then again, I’ve never bought the argument that painters need to mix their own colors for a painting to be theirs. Having worked various layers of the web rendering stack, it also never made a lot of sense to me that you must use a certain layer to be considered “low enough”. Whenever someone claims that it’s JavaScript in Three.js, I immediately suspect, maybe unfairly, that the person has only the faintest idea of how many systems turn their loose instructions and turned them into pixels on their screen.
I don’t doubt that there is technical feat and prowess in being able to do it, and if this satisfies your vision, then you were able to contract an AI agent to do in full, but it is still a layer away from the personal effects of art done by one’s self. I just feel that this still leans more towards some new type of commission.
But largely more is this is a display of how we as people view what AI _seems_ capable of, and how we aren’t being fully aware of what actually has been accomplished. I appreciate the willingness to put this out for the world to see, criticism or not, yet if this is a reflection of how you want your website to be perceived, then having it in a ready, working state tells me how carefully you coordinated the effort. Right now, it just feels reinforces the narrative that work with AI has lost a large sense of quality, even if in this case, it didn’t handle that part. I suppose I expected the article to be something different, but that is on me.
> One of the foremost tasks of art has always been the creation of a demand which could be fully satisfied only later. The history of every art form shows critical epochs in which a certain art form aspires to effects which could be fully obtained only with a changed technical standard, that is to say, in a new art form. The extravagances and crudities of art which thus appear, particularly in the so-called decadent epochs, actually arise from the nucleus of its richest historical energies. In recent years, such barbarisms were abundant in Dadaism. It is only now that its impulse becomes discernible: Dadaism attempted to create by pictorial—and literary—means the effects which the public today seeks in the film.
> Every fundamentally new, pioneering creation of demands will carry beyond its goal.
Walter Benjamin, The Work of Art in the Age of Mechanical Reproduction, 1935 https://web.mit.edu/allanmc/www/benjamin.pdf
Edit: to put a finer point on it, AI art like this is, through it's rigorous application in almost every previous medium, showing that there is a demand and energy for a new art form that depends on the changing technical standard and capabilities that AI has introduced. I don't know what it will be, but you can see some of the edges of it with how it is making waves in all the other mediums.
Interesting write up, the trick to diff renders to have agents check for regression during optimization is clever. The whole thing VERY much scans as an ad to me though. I like company tech demos but I'm not huge on reading to the end of a post talking about how awesome anthropic's walled garden is to read that he's an employee.
Also, the actual website fully crashed Firefox on my phone after making it play horrible grinding sounds over my headphones for like 30 seconds. That's a new one, even for heavy vibe code.
Great article, though if the author ever reads this: please choose "artificial", next time! It's already the first half of AI. If I'm bound to be an irrelevant spec in an irrelevant backwater of an irrelevant galaxy in the moments between abiogenesis and abioterminus, I can at least hope to make one obvious adjective a "thing".
That one Anthropic engineer made "vibecoding" happen overnight, so I write this with full confidence that they have magical influence over the ever-forming neologisms.
One small correction: Andrej Karpathy coined the term "vibe coding" in February 2025, a year after leaving OpenAI for the second time. He joined Anthropic in May 2026.
The euphemism treadmill never stops. Even now, the liar-in-chief is pushing "super-intelligence"
Purveyors of slop will go to any length to rationalize concealing the origins and nature of their wares.
Idk, “artificial” seems about as honest as it gets! And what you’re describing so pejoratively is in large part just how languages evolve.
> I started by creating a new Claude Project. Projects are becoming a layer that spans Chat, Cowork, and Code; you can think of them as sitting above domain-specific concepts. You talk to one Claude with the ability to spin off additional sessions, share files and memory between them, and supervise the entire fleet of Claudes to achieve goals.
> Earlier this year, I would have joked that I get a lot done by telling a bunch of Claudes to do work. Now, I’ve put one Claude in charge of each of my projects, and in turn, that Claude knows how to manage tasks, investigations, and threads. We’re talking more about goals and less about how to achieve them. My little Project setup.
> Every thread gets its own computer in the cloud, where it installs whatever it needs (Blender, FFmpeg, Playwright), checks its work with screenshots from a headless browser, and publishes a private preview page for me to click through. > All threads share one folder of files and one memory. Claude is smart enough to write down decisions I made in memory and makes sure newly started threads follow them without me having to repeat myself.
I must be missing the setting in my claude/claude code - how does one do that?
A project is just a custom prompt and some files you can attach.
Claude code is the same thing with some .claude files, and the folder directory is local to your machine.
Where is the ON switch for all of this? Showing the method and steps and prompts would be more useful than this post.
It appears that they're replacing the "old" version of projects you describe, with a new version mentioned in the post. It's being rolled out in stages:
https://support.claude.com/en/articles/9517075-what-are-proj...
[flagged]
It's all getting abstracted away.Even our computers are.But it's cool to see people continue to pushing the envelope.With the latest models releases, I'm starting to think about what even are humans for now.I guess we will continue evolving and keeping track but with how the models are competent now, I can already be satisfied for the foreseeable future.I don't think I would need anything more.And yet we will continue to push.
Besides the About page and this one blog post, all other links on the site are broken. There’s no need to feel defeat.
Super cool, love the aesthetic. Buy testing the whole blog a bit better would've surfaced the "all posts" link going to an empty page I guess? Also: not reading code for a toy project like this vs in a commercial setting shouldn't be compared at all. Would also really love to see the token spend on this.
I appreciate sharing the process and I totally get this workflow style, I wish you had also shared your usage cost building this.
So Felix is responsible for all that font-mono 01, 02, 03 out there?
There's a tangible quality to mechanical creations. Miss that hands-on feel when debugging distributed systems all day.
What did it cost, everything included?
Yes! That is what I came to say: it's interesting, but I can't contextualize the workflow he describes without knowing the the quantity of tokens involved and their retail cost.
[flagged]
[dead]
I loved the finished work. I'm looking forward to more art in the world.
Agents allow you to produce more, even the silly ideas you can't put too much effort into.
I really loved the Herzog parody, the overall taste and the parts I explored (I'm sure there's more). Well done!
What I really look forward to is people on IMDB creating alternative movie endings or alternative scripts instead of bashing the movies :)
Or in addition to, because the movie bashing is quite funny in itself.
I'd love that too, but I don't see that happening.
Fan fiction is not easy legally [0]. I expect if it happens it will be found on much more underground places than IMDb.
[0] https://en.wikipedia.org/wiki/Legal_issues_with_fan_fiction
Well, they could just share their prompts :)