Brandon Thomas, developer of ArtCraft, a suite of open-source creative tools meant to replicate Adobe applications for free, has now announced apps to replace Microsoft Office tools such as Excel, PowerPoint, and Word, respectively, named GridCraft, DeckCraft, and WordCraft, that are created with Claude. Thomas has shared his new clones to ArtCraft GitHub, making them free for all to download.
You’re gonna have to get over it.
This guy and his crew speedran twenty years of FOSS development over a week of manic effort. What the fuck have you created lately?
I dug up dirt with my hands and filtered it into clay and made a little bowl.
ai can absolutely get fucked.
Anyone script kiddie can fork libre office and call it their own.
So I guess I just did it 3 times while writing this comment, idk what took this dev so long
These are clean-room re-implementations.
Do you think LibreOffice includes a 99% identical Photoshop clone?
Clean room implementation would be taking the spec, API calls, etc. & writing the code that implements them from scratch. This is de-compiling existing code & converting it to another programming language. Same source routines in another language.
It’s like taking a book written in Chinese, translating it to English, and calling it your own.
Bullshit.
Which they did.
This is not a decomp. They gave the tool all of the available specs, and the expected behavior, and let it run. They had a suite of test PSD files but no Adobe software.
Please explain to these other commenters how and why that counts.
May I ask where this is documented? I tried searching, but there’s just too many results to parse, and/or my Google-fu is failing me.
I find it hard to believe near full feature parity could be achieved in under a month for multiple major apps even for an experienced dev using AI. It’s a bit hard to buy.
This project is weeks old. I dunno what detailed writeup you expect to exist. But the project’s Github page says: “Implemented from public specs and observed behaviour only. No proprietary code, shaders or assets.”
There’s no secret sauce in Photoshop. Image editors have been done. Every individual feature is a tractable problem. You can now buy your way out of those, in bulk.
Which brings us full circle to not original code, but I get that’s par for the course with AI. However, I can’t see completing all those apps - with full compatibility - in that time span with nothing but documented specs (especially nowadays when full documentation has long become an afterthought that never becomes reality far too often).
Sorry, not buying that there wasn’t decompiling involved. Dude needs to prove his “work” by releasing every detail of how he accomplished it so others can replicate it before I’ll eat my words. And he needs to do so now - so current tools, hardware, etc can be validated by Joe Public.
Argumentum ad nuh uh is somehow worse than all the sea lions saying your definition wouldn’t count even if you believed it.
C-clean room AI?
So like someone wrote trillions of lines of training data just for this project?
Ooorrrr did a fancy search engine hype them up that they’re a cool kid now and they wanted to believe it so bad they’re willing to claim ownership over anything
No such thing as clean room LLM output, unless you wrote everything in its training model.
It’s in fact by definition as dirty as possible.
Right, because clean room means you’ve never seen any code in your life.
People - spicy autocomplete was not going to derive C from first principles. Neural networks are example-based. The model still obviously never saw Photoshop’s source code, because that shit’s not on Github.
I’m with you on your first paragraph - unless we’re doing some pioneering work, we learn it from somewhere.
But the possibility of decompiling means you don’t have to have access to the source just so long as you can access the code is some form - including the generated executable.
The models were built upon the data, code, art, and texts stolen from a generation of internet users and artists, and from generations of writers.
Do you seriously believe that these companies which built themselves on theft would stop stealing once their first model was built?
If a single adobe programmer (or even employee, depending on how lax their internal security is) used a cloud-hosted LLM at work, you can be assured that whatever code and data that malware could access from their computer ended up being part of the next generation of that company’s model, and eventually of any other’s.
Even if they didn’t, if any windows 11 computer had access to the code you can be certain it scrapped it and sent it home to be given to openAI and any other “AI” company willing to pay.
Thieves will steal. No LLM model ever started clean, and they only get filthier and filthier with time.
That’s still not what clean room means.
Y’all will bend over inside-out to pretend these models need a terabyte of linear algebra just to copy-paste a text file they obviously did not have.
One thing that really puts me off AI is how its boosters talk to other people.
Also I disagree with that label. I’m taking the radical position that… it’s probably fine. It largely works, and people can do stuff with it. Staying mad about it is poorly justified and bad for your health. It’s just a chatbot. Make your peace.
Meanwhile the insistence that ‘AI [thing] isn’t [thing]!’ gets shakier and fuzzier, but no less loudly confident. An AI cartoon isn’t a drawing, sure. But it is an image conveying someone’s idea. AI music is still a song, by any definition. It’s usually well within western music theory. AI reverse engineering has demonstrably produced byte-for-byte identical results, and some folks still insist it doesn’t count, somehow. What is anyone supposed to say to that? It’s an absurd denial framed like philosophy. A counter-poke and saying ‘relax’ do not make someone a delusional cultist.
You have to be able to understand disagreement in terms besides an opposite extreme.
God the reply is SO RUDE.
Neoliberalism sure has poisoned our minds.
Despite what bosses and market pundits will tell you when they’re trying to underpay you, the value in a software development effort is not just the end product that you can sell to customers — it’s also the knowledge of how the codebase works, and how to modify it over time.
I’m not an ostrich when it comes to AI-powered development. I’ve been to multiple conferences and courses over the past year specifically about developing with AI. My Youtube and Substack and podcast history is multiple hours per day of content about AI news, practices, and opinions. My weekly visit to my bookstore has been mostly to pick up or place an order for books about AI. I use agents for my daily coding tasks at work. I have multiple local LLMs on multiple machines at home that I use from multiple harnesses. I have ralph loops, I have full-AFK workflows, I have custom sandboxes.
I could go on, but you get the point: I’m not standing on the sidelines when I say…
The vast majority of AI use today is horrifically irresponsible. Current AI use is digital asbestos in the walls of our technological society. It will take generations to fully recover from this. I am not excited, I’m not gonna “get over it”, and I’m not gonna grant any degree of false equivalence between actual FOSS and these slop stunts.
The bubble cannot pop soon enough. Not just so that we can minimize the economic damage, or so that hardware prices can once again experience gravity, but so that we can actually be honest about what this stuff is and how to use it.
almost everyone wants to get a good deal, corperate people included.
If only some technology could could summarize huge quantities of text, or modify code the user doesn’t fully understand.
Bemoaning capitalism is certainly one response to an aggressively libre / gratis clone of infamously greedy software.
The bubble will pop any day now. God willing, every leading company swiftly goes under. And then, people will continue using this technology, because long-term maintainability is such an absurdly high-level concern that it obviously won’t stop anyone from exploiting this witchcraft. This is just a thing computers can do now. It will continue to improve as people fuck with how networks are designed and trained, even if none of them are throwing around three-comma budgets.
Businesses won’t need you to understand and support their processes, because they’re not gonna hire you, they’ll just make the software themselves. It’ll be a fragile mess, much like their databases. And much like their databases, so long as billing works and they keep making spoons or whatever, apparently that’s good enough.
Downvotes won’t change that this works.
Sneering at people using the tool won’t make the tool go away. It produces running code, to spec, real quick.
I have a few original projects I haven’t been arsed to finish (or in some cases, start) and now I could have them finished by close of business Friday, for the cost of a night out. And y’all would make excuses for how it somehow doesn’t count. It’s not my idea, it’s not my intent, it’s not my design, it’s not my insight… whatever. It would make something real which doesn’t exist yet. I would be the only human to blame for it. Once it is demonstrably made, either I made it, or the robot did. If your philosophy can’t handle that then your philosophy is wrong.
If this project still exists in a year, I’ll eat my words
It’s already in a functional state. Eat now, braggart. If they achieve their goals within the month, clap the dust from their hands, and walk away with the job done, how is that less impressive than bumbling along for an entire year, to accomplish the same thing?
Cause it was scraped from those original projects and doesn’t exist without them
You can flex it “only” took the entire dev time of libre office, all open source software, and every book that has been scraped + the time it took the dev to make, but to argue they made this solo is laughable.
If you want to sit with the big kids you gotta be able to handle the heat, Microsoft changes their document format more than their taskbar - I strongly doubt this dev even understands how their files are read.
Do you think there’s a Libre Modern Warfare 2? Because the same tool did the same shit to Modern Warfare 2. Show me the open-source repositories that simply copied from.
I do believe there are more shooters than just MW2, is it at all binary compatible? That would impress me (cause its impossible for an ai without scraping the source lmao)
Its like how they ‘revived’ the dire wolf by getting a white husky. Like imitation isnt equivalence, even if you slap the same title on it.
Goalposts moved from ‘it’s only copying this exact project’ to ‘any conceptually similar project can be reworked until it’s identical and that doesn’t count either.’
The MW2 decomps are from the binary - but there’s obviously none of its source code in the training data. Not that you seem to care, suddenly. You’ve made your position unfalsifiable and therefore meaningless.
New clone of a whole-ass software suite? They must be copy-pasting a specific libre project, doesn’t count. Oh it’s from spec documents, and structured completely differently? Well the other libre implementation is in the training data, doesn’t count. Oh it also cloned cloud applications with no binaries or alternatives? I do believe there are more video effects packages, does! not! count!
Meanwhile.
These schmucks have a one-to-one copy of the latest Photoshop, permissively licensed, ready to use. It did not exist last month.
Whining about how it got made won’t change a goddamn thing.
Apparently you don’t understand what “maintenance” is.
This slop project is going to fall apart within a year due to being utterly unmaintainable. Will it take Adobe and Microsoft down with them? One could hope, but that doesn’t make their codebase good.