Mea Culpa – Dark Hours(blog.terrygodier.com) |
Mea Culpa – Dark Hours(blog.terrygodier.com) |
... but I'm not going to insert a link to his project - https://github.com/mbeher2200/DarkHours
I don't think that's a bad thing, or that it's necessarily plagiarism - unless the LLM spews out the other project's code from its training...
I'm happy to see multiple takes on the same idea. There are lots of tiny aesthetic and functional decisions that make apps different from one another. Now that implementation is cheaper we should celebrate that.
“It became clear to me that the web app I had launched using Claude was strikingly similar to his open source project, even reproducing a bug he had later fixed.”
At what point do we confront the reality that premier AI models are blatant copyright violation machines through and through?
> While I had genuinely never seen DarkHours.app before yesterday
If only there were a way to share your own commits resulting from your own Claude sessions to verify whether it's true...
The AI eating his homework wasn't even the worst of it, and this whole debacle is just a series of poor decisions and deceptions. Bizarre. Yet there are comments in here praising his "attitude". Can I interest you fine folks in the Brooklyn Bridge? I'm selling it for 40% off, today only. Message me.
This is easily the most casually bizarre HN saga that I've experienced...
He fucked up and is owning up to it. Sure we all wish it never happened in the first place, but since he did, we prefer he confess rather than double down and go on offense or something else. The AI didn't "eat his homework". He's not a student, this wasn't a homework assignment.
No? He doubled down hard and proceeded to insult all the readers' intelligence. Just in case it's not clear: no one believes that Claude accidentally created that bug-to-bug ripoff app.
Read the link and the previous one: https://daringfireball.net/2026/08/retraction_app_store_reje...
Open source isn't just code, it's an idea, a public square, and a community.
Fork a repo and customize to your heart's content? Go for it!
Fork a repo, change nothing substantial, and publish it with a nearly identical name in the app store and on a nearly identical domain? At the very least, that's in poor taste.
> I fully support open-source, but I support way more the creativity and idea from individuals that can generate something new.
None of that happened here...
There is some copyright discussion on one of the other submissions on this topic.
The developer's astrology app containing tarot reading features was rejected from Apple's App Store that prohibits astrology apps.
He then replaced the app's content with a clone of, if not the identical open source astronomy app "Dark Hours", copying even the name. The original is here: https://darkhours.app
The developer was then in contact with John Gruber, who wrote an article claiming Apple's reviewers confuse a nice, scientific astronomy app with astrology. The developer reviewed the article before publication.
It then turned out that the developer forgot to mention that originally he did upload an astrology app for review. Gruber felt duped, and noted that it's unclear whether an updated build containing the new astronomy app was ever submitted to Apple for review: https://daringfireball.net/2026/08/retraction_app_store_reje...
Now the developer seems to claim that he merely accidentally cloned that open source app when working with Claude.
I am not buying it.
Coincidentally:
Just in the last few weeks, D'addario, one of largest guitar string manufacturers, posted on Instagram what is clearly an AI-song to advertise its new line of guitar strings.
After being called out on it being AI, they posted a mobile-recording of the song recreated in a DAW, but it didn't even match the timing or the notes played.
After being called on it AGAIN, they blamed it on LANDR and Apple Mastering "changing the notes", but those only do "AI knob tweaking", they're not generative-ai audio synthesis. LANDR itself replied.
After being called on it AGAIN, they finally caved and said one of their workers used Suno without their knowledge.
After someone contacted the worker, he said he was pretty clear in his communication with D'addario that he had used Suno.
EDIT: The worker claimed it was "his first time using AI" and "he didn't know", but after someone did some digging in his Instagram profile, there's a few Suno songs and a lot of AI posters and avatars.
This sort of ass-covering is getting crazy with AI...
There's no way Claude/AI recreated another persons app exactly, with the same name and branding, without being told to do so in some way. I just don't believe it.
And him not using it and calling out AI is just him covering his ass
I refuse to believe nearly 100% of game developers aren't using AI, but how many put the tag of shame on the Steam store page?
It's a very strange situation.
The irony is how unnecessary it was. They could've just said, yeah you're right, we used AI. This is a level of lying you reserve for covering up a murder or something. Maybe they're just worried (and they're right to be worried) that it's easier to make music with AI than with their strings.
The simple truth of the matter is that contempt for your customers is now normal. I expect this from Fender, now; they are utterly without soul.
But D'Addario are really one of the most analogue companies ever and they are still family-owned; they should have understood the magnitude of this problem right at the beginning and just apologised.
At least one of my three guitars will likely continue to have their PB strings, probably, because they still sound the best on my cheap Weissenborn clone.
But it is very disappointing and it leaves a bad taste. Perhaps I will revisit Rotosounds (wound in a factory very near where I grew up) for my PBs. The UK also has Newtone strings so I really have no reason to continue to support D'Addario (though the low-tension Aloha set I really like for the Weissenborn copy that lets me have a 15 on the high D is always backordered).
This is not the defense they think it is.. If your mechanic rotates your tires and improperly torques the lug bolts "it was my first time using a torque wrench" isn't excusable. It's good to experiment and learn new things. It's not good to throw your little self-study project out into the world with your employer's name behind it and present it as production worthy material.
I think it’s shaping up like he didn’t “forget” to mention it. This looks like an intentional play for attention. It worked very well. Too well, in fact, which drew enough eyes on to the situation to start noticing the cracks in his story.
A key part of the Gruber retraction article is that Apple rejected a different app named Asterly. He then copied the Dark Hours app, including the name, from someone else and tried to imply that was the app at the center of the controversy.
If he had actually written his own astronomy app and swapped it into the story he may have gotten away with it. Doing the wholesale copying of someone else’s work, including the name of the project and stripping the license in the process, is what got him caught.
App Store rejection stories are always popular. I see at least one post a month on Hacker News from an app developer trying to get attention because their app was rejected and it usually receives moderate social media attention. I think this person saw an opportunity to ride that anger for some personal publicity and went for it. He flew too close to the sun and got too much attention, though, so his mistruths were revealed. Now he’s trying some CYA techniques because the original ruse backfired.
Are there any links to this stuff that you can share?
[UPDATED] Ah. I see this post has it[0].
The subsequent blog post heaping it on Apple and the second app that was just a clone (and wasn’t what was rejected by the app store), the HN drama… that stuff all just sounds like everything snowballed from that first made-up story in a chat.
I, too, feel entirely duped. What the heck dude. Uploading an astrology app that later pivoted to astronomy and having that be rejected for astrology is a way different problem from just having the latter rejected for astrology.
Thanks for drawing the chain. Infuriating.
You are misattributing a quote.
https://daringfireball.net/2026/08/retraction_app_store_reje...
And an identical bug, per the Bluesky convo Gruber linked.
https://bsky.app/profile/mmmeh.bsky.social/post/3mslc3b3u4c2...
There have been quite a few copyright cases alleging this behavior that LLMs are not substantially transformative[1]. Note that cases have largely been ruled in the favor of the AI companies, with courts ruling that model output is “substantially transformative” of training material (although what is currently legally permissible may not be what we feel is socially permissible or ethical). The the fact that these allegations continue to be made, and the fact that AI companies seem to be trying to cover their asses by disallowing prompts which would be likely to create content highly similar to that of copyrighted training material [2] might be enough to at least warrant some consideration that Godier is being honest here.
Not that this would absolve Godier (quite the opposite in my mind- low effort ai workflows produce slop, and he chose this intentionally), but perhaps some of the onerousness here also falls on the AI companies pillaging of work.
[1] https://copyrightalliance.org/ai-copyright-lawsuit-developme... E.g. https://nytco-assets.nytimes.com/2023/12/NYT_Complaint_Dec20...
[2] https://arstechnica.com/ai/2026/07/chatgpt-stops-cloning-fam...
That seems to be the thing everyone is into, these days.
(from All You Zombies, by Robert A. Heineken, 2026, ebook)
(a damage-control strategy where an organization or individual admits to a portion of a secret or scandal after a cover story fails, while still hiding the most damaging key facts.)
An LLM/agent harness generally doesn’t just claim another project wholesale, including specific bugs, as the base of your idea, unless you ask it to.
Now that you've introduced the phrase I can see it in a lot of recent scandals.
Peter Attia, a popular health and longevity podcaster who was revealed to have substantial ties to Jeffrey Epstein used this exact trick recently. When the Epstein e-mails were released it was discovered that he continued to consort with Epstein after all of the revelations. Even worse, he talked to Epstein about the fallout from the accusations and continued to cozy up to him. He was trying to run a very expensive concierge medical clinic and relied on Epstein's connections to get more clients, so he wanted to continue keeping that relationship warm.
When the e-mails came out he issued a public non-apology that implied he had no idea about any of Epstein's activities. It was plausible enough if you never saw the actual e-mails and thought Epstein was just another contact in his network, but the actual e-mails revealed that he was continuing to converse with Epstein after the revelations, including about the revelations!
So "limited hangout" is precisely what he and his PR team tried to execute. It seems to have worked on a limited set of his fan base who still try to downplay his Epstein connections to this day. The e-mails are easy to find, though.
>> The words are non-intuitive
Yeah.
"Incomplete mea culpa", "Incomplete admission", "Incomplete penance", "Incomplete contrition", "Incomplete remorse", actually say "the thing" instead of using a new random phrase.
I like "incomplete contrition".
Did Godier copy the other guy’s repo, name and all, and try to pass it off as his own work? That seems like the obvious conclusion, albeit brazen to the point of insanity.
Did Claude copy the repo, in response to a prompt? If Claude copies a Claude vibe-coded project, can any plagiarism really be said to have taken place?
Or is this just what Claude is statistically likely to spit out if you prompt it to build an astronomy app and leave it to it?
You can trace back to the original repo from that thread. The key points are that the copy had
1. The same name. One was "Dark Hours" and the other was "DarkHours".
2. The same highly specific bug. I didn't investigate closely but from the description the bug sounds too unique to have been identically created in a fresh Claude-style duplication.
The duplicated name is the most damning part to me. I've never tried to copy a project with Claude, but I can't imagine the sequence of events that would lead to Claude copying the name of a repo but going to lengths to disguise other things like stripping the license. This all seems highly intentional, like someone thought it would never be noticed.
A name clash is the least damning evidence in my opinion. Last year my previous employer asked us for suggestions for naming an AI agent. Everyone produced pretty much the same list, because everyone just asked ChatGPT.
Except me. I had also asked ChatGPT, but I bothered to google the suggestions and found there were already like half a dozen AI agents for every suggestion.
(they ultimately launched what was by then the 7th or 8th AI-related product with essentially the same name)
Creativity in naming doesn’t seem to be their strongest suit.
https://blog.terrygodier.com/2026/05/24/ive-been-a-little-di...
This was a beautiful ending
It's as if real people are not all good or all bad, but can do good things, and bad things too.
Published a rip-off? "Sorry, Claude must have done this".
Published a shitty/wrong/insensitive/<insert other adjectives here> post somewhere? "Sorry, the AI wrote it"
Regardless of how we view LLMs as a whole, if you produce content with AI and publish it, you personally own every line of what you published. "Claude did it" cannot ever be a valid excuse.
I think this an incredibly important point. As soon as we devolve reposnsiblity to computer progams, society cedes control to the algorithm vendors.
Jeff Geerling was called out today for his clickbaity video entitled "Intel just matched Apple Silicon. Seriously.". Jeff's personal response was "I had three YouTube titles, this one got about 2x the watch time as the other two, so I let YT pick it."
Another commenter replied: "You are trying to launder responsibility for a clickbait title to a computer program." I think this exactly nailed the problem.
"I was genuinely only using Claude to make my writing clearer/redraft/to organize my thoughts because I can't write as well as you can"
The new one for me this week was someone claiming they write exactly like Claude does.
I find this synopsis very hard to believe, seems he's shifting blame onto an LLM instead of taking responsibility for his actions. He said he'd significantly alter the feature set, so why didn't that happen? That's a single prompt to Claude. He more than likely just told the LLM to clone the repo and ship it as is, not actually working on it in any capacity. In that case it wasn't the AI "reproducing" a bug as there was no reproduction in the first place.
That's a terrible lesson.
It's an open source app. Tell you agent to fork it an contribute to it if you have any ideas that might improve it.
You don't want to build castles anymore if you don't own the whole sandbox?
> To the best of my recollection, this is the first post I’ve retracted in the 24 years I’ve been writing Daring Fireball. I hope it’s the last. I was misled, both overtly and through omissions, in several ways, but what I publish is my responsibility, and I apologize for the error.
It's disheartening for those of us who intrinsically value human authorship.
I think to some degree this is going to be inevitable for a while. No matter what people think about the ethics of AI or AI use, the fact of the matter is it exists, and it is shaking up and going to shake up a lot of the world. I personally think "AI" isn't really a product anymore than "The Internet" is a product. It's infrastructure in the same way the internet is, or to quote "Halt and Catch Fire", it's "the thing that gets us to The Thing".
The problem is no one really knows what The Thing looks like, so everyone is throwing anything they can think of at the wall and hoping to see what sticks. And they're dong this because they remember Sears and they remember Blockbuster. No one wants to be the next company that was positioned perfectly to integrate the new technology into their business model and missed the boat completely for short sightedness or lack of speed. "$foo but with AI" is the current generation's "$foo but on the internet", and everyone is scrambling hard to not be the next company that someone writes a "whatever happened to them" youtube video about.
Incidentally, this "AI as infrastructure" view I have is also why I really think governments should be getting into the space of building and distributing training data sets and models. Not only because it could reduce the overall amount of bot scraping and traffic if everyone who wants to build could go to a known good and clean source of training data, but also because nominal usage fees could help manage the copyright concerns. I've pitched it before in more detail, but I imagine something akin to radio broadcast royalty licenses, where people whose copyrighted works are part of the government training sets are paid out "royalties" from the pool of money brought in from licensing the training sets.
Except that some portion of D'addario's customers--presumably those that actually buy their guitar strings to create music via real guitars--aren't okay with it.
Just blame the next thing down the chain that can't defend itself, so you don't have to take responsibility.