This idea reads like a joke, but there's something to it.
One feature request: In addition to high-level milestones, it would be cool if a partially-funded project would generate a public, highly detailed implementation plan.
Also, IANAL but MIT is still a license with a copyright holder. I don't think saying "it's MIT, we all own it" is defensible. The courts might view all this code as public domain.
I wonder why people are more eager to pool money to pay a corporate-owned computer to build things than to the actual humans who have been building open source for decades? Much of which has ended up in the training set?
Fablepool spends Anthropic's inference budget and puts the output under MIT, thats not growing Anthropic's moat, its commoditizing it. Anthropic just sells the API calls either way.
LLM output is just very predictable. If you spend 200$ on that /goal you will get some output. It may or may not work perfectly, but there will be a repo with some progress. If you specced the goal well and it is feasible, a decent model will most likely get you decent progress.
Also who would take on any of these projects for a meager 200$? Most of that stuff is borderline interesting, clearly not interesting enough for the people proposing the things to start working on them themselves.
I think what will be interesting is not whether the code will be produced, but rather: will anybody actually use any this output?
This sort of reminds me of startups that go out of business and then open source their code. It's kind of cool when they can do that, but almost nobody ever gets value from it.
Anyway, if anyone uses the code produced this way in prod, I'd love to hear your story.
Ownership is not the same as authorship. Copyright is about authors rights, not owners.
If you buy a text of me, I cannot sign away my authorship, and there’s certain limitations on what you can do with my text regardless of contract. I can only sell you usage rights - which may or may not be exclusive. If the text I wrote is trivial, neither you nor me can limit when it is reproduced. The effort of collecting data is not sufficient, if the data itself is declared trivial. See rulings about phone books.
When an AI provider produces data that is deemed not copyrightable, it cannot legally sell you exclusive usage rights. It can give it to you exclusively, but since you cannot yourself claim copyright, the moment you publish it it becomes available for others to use as well. One may argue that an LLM is similar to a phone book, with its entries being “trivial“ and its composition not artistic enough.
"I paid money to someone to write that code" is exactly the line of argument that would lead to no copyright. There was the famous monkey selfie court case in 2015 that ruled that at least in the US, monkeys can't hold copyright. The same arguments would apply to AI. And since the AI can't create copyrighted works on its own, it can't assign copyright for works it created for hire either
The other line of argument is the "Claude Code is to coding like a photo camera is to painting". The image is generated automatically, but the input in how you point the camera is enough to still make it a creative work protected by copyright. Under that interpretation, you are not hiring AI, you are using it like a tool
The US Copyright Office holds the former opinion. I'm sure once this goes to court, lots of companies will vehemently argue the latter. I would not be surprised if we even end up changing the law over this
>"I paid money to someone to write that code" is exactly the line of argument that would lead to no copyright.
That's news to me. I (along with many hundreds of others) was paid to develop Minecraft, candy crush and battlefield, yet last I checked, they all retain their copyright.
Because you are human, and can thus create copyrighted works, and can assign that copyright when you do work for hire. Monkeys and AIs can't create copyrighted works, so there is nothing to assign when doing it for hire
The other line of argument avoids that issue by arguing that you personally created the code with the help of a tool (like a compiler or camera), not just commissioned it
GP's (or rather GGGP's) statement was "I did not write that code but I paid money". So they claim no authorship. Anthropic is a company, they can't be the author either. So the only one left as author in that reasoning is Claude.
I don't think that necessarily anthropomorphizes it. We speak of monkeys as authors without calling them human. And really the legally important fact is that there was no human author. You can also treat it like CCTV footage which is generally not under copyright because there is no human author (even though most would hesitate to call the camera the author either)
As a human, it is possible for you to create copyrightable works and transfer the copyright to Microsoft in exchange for money. It is not possible for Claude Code to do those things because Claude Code is not a human.
They retain ownership over that code because you signed a contract saying explicitly that. Did you sign the same thing with Anthropic? Did your company?
Anthropic isn't a party to such contract either because they (most likely, in the reasonable reading of relevant laws) hold no copyright over the output of their LLM.
In this case, it's more like you paid money to FablePool. FablePool used Claude as a tool, and it delivers the product so they are the owners of the code, and have the MIT copyright assigned to them.
Nope. Copyright simply doesn't work like that. Unless there's real human creativity involved in the process, they can't just claim copyright to LLM output.
I don’t know, if the design itself is copyrighted you could argue that the AI is just a bunch of hired workers that built it for extremely low wages.
If I hired a bunch of people to build me a house, and I drafted the architectural plans with the help of a paid architect, neither the architect nor the builders have ownership over the home.
So if a collection of people design something together maybe that has merit, they collectively paid for Anthropic to build it for them…
I’m pretty sure copyright office has settled that already. Inly human expression can be copyrighted:
> As described above, in many circumstances these outputs will be copyrightable in whole or in part—where AI is used as a tool, and where a human has been able
to determine the expressive elements they contain. Prompts alone, however, at this stage are unlikely to satisfy those requirements.
I think pooling/donating tokens will be a thing. Not sure if like this, but in some format. The Django project, for example, came out and said they don't want your tokens, but I think a lot of people/projects will (do?) want your tokens.
I guess if you have a subscription with a token allowance that you are not going to use up this week, it’s better to let someone else use those tokens rather than throwing them away. So using the food analogy it’s more like a store giving away unsold sandwiches to the homeless at the end of the day instead of throwing them away.
Donating tokens to a software project is a bit like donating food to a hungry person.
I think it might be beneficial to use blockchain, so that the donor can audit which prompts the token-pool they donated too performed. Perhaps donating tokens can also give you votes on which prompts are entered.
It’s a lot more like giving a hungry Hindu a gift card to a specific non-veg restaurant. Maybe they’ll use and enjoy it, maybe they’re vegetarian and will be insulted; either way that restaurant benefits. Especially if the hungry person exceeds the value of the gift card.
The good ones all seem to be pointing in the direction of Django. Which, on its own, says a lot about how likely people will care about vibe-coded anything, whether pooled or not.
This is exactly how I'm building an OS right now. I have a lot of things speced out, and for most of them also create an issue. And I have a friend that just points his claudr code at the repo and tells it to "find the next thing to work on and implement it"
I then do the review, verification, etc, but a great way to used unused quota.
I've always wanted to figure out how to implement a cooperative source license. Something like, you're allowed to do what you want with it, but any derivative work requires the same license, and X% of any income goes to the cooperative?
Not sure how it'd work, but there's absolutely a niche for a privacy focused data cooperative out there.
Any income from what? The code is free, right? X% of your company's total revenue? Might as well just say "companies can't use this".
Personally I like the idea of a "free as in freedom but not free as in beer" license. You have to pay for a copy of the software, but after that you're free to use and modify it as you please, and share/sell your modifications under the same license.
To turn that into a cooperative you could have a company own the code and pay developers in shares of the company for PRs or other contributioins.
Should it? If it was real world infrastructure, like a bridge it'd be easier to say that it belongs to those who lead the project and those who put down the money
The nice thing about a CC0 work is that it belongs to everybody. The leaders of the project have the same rights to use and modify the software as they do with software they have exclusive copyright over. In fact copyright does not grant the rights holder any new rights they did not have, it only restricts the rights of other people.
The problem with running open source code is the security aspect, but with Mythos running point, how would you distribute revenue is the real question.
Which market is even left after since the sasspocaloypse?
Rather, it did work at milestone 14, but then regressed at milestone 15, where it changed the link from a wikimedia image to a nonexistent file in /assets (despite still having the "Photo via Wikimedia Commons" caption).
If you check "DEPLOYMENT.md," there is a lengthy list of deployment instructions for the app, and it includes creating an assets folder and putting an image of Claude Shannon in it. There are also other instructions, like "please make a favicon." So I think that bit is valid, the AI is simply farming out work to the human agent.
My question, though, is why the "Live, public build log" only showing up to milestone 3, but the artifacts go up to milestone 15? And there are different index.html pages in the artifacts list, one for milestone 14 and one for milestone 15? Are there different conceptions of "milestone" in here? What's up with that?
I have found that on long-running tasks, many of the communication (and other) invariants get dropped at seemingly arbitrary points along the way. It probably just stopped doing the log.
yeah.. deployment.md had instructions to stick a photo there, but rather than explain I just got rid.
gonna work on a few examples and fund them so people can see it actually work
It seems highly suspicious that all the target cost estimates are for $150-$400, regardless if it’s a bench for pelican on bicycle, or a clone of AAA game.
It reminds me of scam eshops where everything cost $random dollars in a hope that someone will enter a credit card number.
Shouldn't we still have people in the loop for selecting/proposing the best implementation (plan)? Vibe coding an entire solution from a prompt still doesn't feel like the optimal way to write software.
At some point you have to say "Is not having it better than having it?" Where's your dude, today, who's gonna code this? If it were gone happen, it wouldve.
From my 10 years in the .net, it seemed C# devs will pretty much do anything to avoid using the right tool for the job or solving the immediate problem at hand.
I think assumption of the gp is that while Fable might be impressive, even Fable would take a bit more (sarcastically meaning a lot more) than $200 of tokens to solve this quite serious problem.
Similar solution worked for ASP back in 1999. ASP/VBS was terrible slow at string building and Response.Write. Build it in the fast code and then output.
It indicates the level of trust people have in the platform, and the combination of the product-platform behavior. If someone with the wherewithal to solve garbage collection for C# for HFT could actually describe why GC in C# was a problem, they wouldn't be asking for $10. But for $10, for something something you're dimly aware of is a problem? I'd throw $10 at some nonsense I read on the Internet.
> What happens when the pool exhausts but the task is not done?
Interesting that this doesn't seem to use blockchains. Arguably it would have been a good use case. OP, could you elaborate on the reasons for the choice (if it was a conscious one at all)?
I'd love to see Anthropic (or someone with mythos access) create a cybersecurity version of this. So that I could create a pool that says "find security concerns in this github repo." Then the report from mythos gets sent to the code/project maintainer and revealed to the public (that paid for it) at the 90 day mark.
For your information, a group of Mythos-approved users at Apple, Google, Microsoft, and several other Project Glasswing partners have already been doing this for the past few months. We just can’t share many details publicly yet.
The target codebase cannot improve beyond the point that the reports are incorrect and a waste of money.
There is also the question of whether humans can waste so much time reviewing AI code that the vulnerability is not patched before it is exploited. Another one is whether when the human is removed from the loop that the codebase becomes more vulnerable in some other ways.
Would love to see non-technical audiences think like this.
For years I've been trying to get estate agents to support an open-source real estate website builder. The pitch is obvious: instead of each agent paying thousands for a bespoke site, pool resources, fund the features you all need, and everyone benefits.
Getting non-technical people to commit to something abstract before it exists is nearly impossible though. Hope a model like FablePool can change that.
A lot of AWS is built on open-source. This is obviously ignoring hardware costs. I don’t know if it is all that ridiculous anymore. These models are very good at wiring together open-source systems. The world is crazy right now…
AWS has over 200 services, so that's a little over $2 per service. Yeah, a lot of it is built on OSS, but there is a ton of it, and there is also a lot of work involved in building the APIs and web UI, and making it scalable , secure, and resilient.
Now, you might be able to make a version of some small subset of aws services that runs works ok for a small scale for with relatively simple needs, for that many tokens, but I don't think that's what they were going for.
I’m on your side in that I would never take a contract to actually do this, but…
If we swapped out the IAM backend for something extremely simple like just private keys (one per allowed service or JWT-style list all services in the key), then we could have something that looks/feels pretty similar. With a 2$ token spend.
Not at all the same but it would look/feel pretty close.
"A grimoire is a textbook of magic and sorcery. Traditionally, it contains instructions for casting spells, performing divination, creating magical objects like talismans, and summoning supernatural entities such as angels or spirits."
It's how they name classes of models, presumably this implies something about the relative quantization / size of model, not about the specific performance. E.g. Fabel 5 will be better than Opus 5, better than Sonnet 5, etc. The 5 is the version number of the particular iteration / training run at this class of model.
Some people have a drinking budget, I have a domain name budget. A couple years back, my friends threw a domain name transfer party, like a secret Santa where you got to surprise someone with a domain name you got for them.
You don't list contact info on your profile page to contact you with, but I'd love someone (other than Fable/a bot) to work on this with, and you're clearly a like minded individual, so reach out
Before putting in money to this small anonymous website, I'd love to hear about the people behind the project. There's a single mention of 'Barras Industries', but not much mention about them online, or what else they've worked on.
They should have called this "WishingWell". I'm wishing them well, but some of these projects are so over the top pie-in-the-sky silly, and funded with $0.25.
Sounds like great idea, but can’t seem to find some key info: which “public ledger” is used here? A blockchain? If so, there would need to be a massive fee overhead. If not, then how are we not relying on trust?
Also, is there some kind of ownership structure based on investment?
I think the bottleneck is testing. I want to build a replacement for Zwift, a virtual gym game for bike trainers and treadmills, but testing it could be difficult without a real person on real hardware. How does the LLM know about the hardware protocols and stuff like that.
Why are there no reverse engineering projects listed? These would make the most sense actually. Or future features of MacOS, Apple hardware designs (as Apple would not be able to patent them then)
Seems similar to open source bounties, which have been tried in the past and never succeeded.
We've seen something like 20+ years of different attempts of voluntary donations to fund open source, and it never worked. Companies barely fund anything voluntarily.
I'm taking the opposite approach with Supported Source (https://supso.org/) which is this: actually force companies to pay to use the project. Sell commercial licenses. Make it mandatory to using your software commercially. This approach works much, much better than voluntary donations.
I think there's a categorical difference between paying for long term maintenance voluntarily vs paying for something to exist. The latter works much better as the value prop is clear and you can scratch an itch. Kickstarter is similar.
I managed to write one that at least didnt had the font and colors (using 4.5)
Yesterday, I prompted Fable to improve the frontend to make it look different from Claude style, gave detailed examples etc. 15 minutes and $32 dollars (!) later (used cursor lol) it gave me the shittiest more claudiest website ever, basically ignoring everything I asked
First, your server is struggling. It took about 20+ seconds to respond just now, FYI.
Second, it's not obvious to me that I can get my money back if something doesn't pan out / get approved by a certain date from the homepage alone. That might make people hesitant to put anything in if they think it might get locked in there forever if the site dies / you take it down / etc.
Not affordable, unless the devs are in somewhere like Vietnam. And there's still no way they can build as fast. And still, at that price point, quality would be highly questionable. So yh this doesn't survive beyond the joke stage.
If users posted ideas, voted on them and then other people built them then that would be the same. But kickstarter is the producer posting an idea for presale
I forgot to fully describe the prompt since I already described it a bit on the title of the submission, which might be a problem. I hope the title of the submission itself is included alongside the prompt when giving instructions to the AI.
This is actually kinda exciting. I threw in an open-source idea I've been playing with, and paid $25. I hope it comes back up soon or I'm going to have to put Fable on building a replacement.
Why do open source collaboration? Why not a single product developer getting crowd paid to add features, solve bugs, using AI. So many businesses will see their moat wiped out.
On the macro level capitalism is winner-takes-all and Musk is the only one seriously playing the game. End game: own everything, including payments, and governments come begging and will protect him from citizen revolt. Supervillain/overlord territory.
I doubt that's legally decided yet. There hasn't been precedent as it hasn't been challenged yet.
I mean Claude will tell you because anthropic made it tell you that, doesn't mean it's true.
GoFundMe and indigogo aren't responsible for the actions of the funded projects either, hence it's unlikely that any judge would decide that the liability would go to the platform if it can show it's doing it's best effort in moderation wrt illegal content
If you mean just throw it together and then don't moderate at all then .. yeah, you'll be held liable. But that's not because of the person paying the prompt, it's because moderating illegal content is the responsibility of the platform provider.
The coding isn't the hard part. It's the people and networking. Facebook's only moat is HOA boards that think private communication behind Facebook groups somehow equates to public messaging a community...
In other words, once people got on it, it was too late.
"Build a completely greenroom, open source AWS" – $700
This is engineering theatre (pun intended).
The amount of hubris here is exceptional, the author doesn't even know that it's "clean room" rather than "green room". What does it even mean to build an open source AWS? There are many open source IaaS/PaaS components. Is the author suggesting any hardware design, because that's a critical component.
The only possible result of this is an AWS fanfic. An art project that looks vaguely like a cloud provider on the surface if you squint, but with zero substance to it.
And this criticism has nothing to do with AI. You'd get the same spending 100x that budget on any engineering team.
At least it recognizes that energy deserves funding, ideally beforehand. Yet it would be harder to sell if a human asked for payment, even if delivery was guaranteed.
This is DOA. Claude will refuse to work on anything there or better just bankrupt the funding members while delivering slop(on purpose to doge their cyber operations )
I could see something like this working if you actually had a assigned human developer(s) to assist the task. There are few interesting tasks that can actually be completed in one (or few) shot and have anything usable.
Neat project idea, but truly ruined by requiring a google sign-in both to submit new projects and to donate to projects. Dead service to me until that's gone.
Remember, Google aids and abets militaries of governments that the UN has found to be committing genocide.
My first choice would be non-SSO. Let me pick a username (or email address) and a password.
Btw, thanks for the response. Not sure why you got downvoted for it, but you have my gratitude for being one of few devs who are sincerely responsive to these types of concerns.
I got downvoted probably for tone which is my bad but what I’m laughing at is this person doesnt seem to understand why people pay for aws, it’s certainly not the laughably bad console or the buggy control plane. it’s the reliability guarantees granted by their massive physical infrastructure that was meant to replace sysadmin’s running racks in a closet and wrangling terrible ansible/chef playbooks.
this literally already exists if you’re willing to maintain your own physical infra, and has for a long time - nothing aws does is that innovative software wise. maybe their managed k8s eliminates a ton of pain, but I dont know. it’s the reliability guarantee + support + not having to maintain physical servers. if youre willing to shirk all that and do it yourself why would you want aws? lol
This idea reads like a joke, but there's something to it.
One feature request: In addition to high-level milestones, it would be cool if a partially-funded project would generate a public, highly detailed implementation plan.
Also, IANAL but MIT is still a license with a copyright holder. I don't think saying "it's MIT, we all own it" is defensible. The courts might view all this code as public domain.
I wonder why people are more eager to pool money to pay a corporate-owned computer to build things than to the actual humans who have been building open source for decades? Much of which has ended up in the training set?
Exactly. This screams crowdfunding for AI labs. Who made this? Someone at Anthropic?
Humans are more expensive.
Humans can also use LLM's even a local LLM
That's even more expensive.
Fablepool spends Anthropic's inference budget and puts the output under MIT, thats not growing Anthropic's moat, its commoditizing it. Anthropic just sells the API calls either way.
LLM output is just very predictable. If you spend 200$ on that /goal you will get some output. It may or may not work perfectly, but there will be a repo with some progress. If you specced the goal well and it is feasible, a decent model will most likely get you decent progress.
Also who would take on any of these projects for a meager 200$? Most of that stuff is borderline interesting, clearly not interesting enough for the people proposing the things to start working on them themselves.
I think what will be interesting is not whether the code will be produced, but rather: will anybody actually use any this output?
This sort of reminds me of startups that go out of business and then open source their code. It's kind of cool when they can do that, but almost nobody ever gets value from it.
Anyway, if anyone uses the code produced this way in prod, I'd love to hear your story.
if fable is writing it, courts my declare that its not even public domain? not a copywrightable work
I don't get this. No I did not write that code but I paid money to eg. Anthropic to buy that code. To me it sounds like I own it just the same.
Ownership is not the same as authorship. Copyright is about authors rights, not owners.
If you buy a text of me, I cannot sign away my authorship, and there’s certain limitations on what you can do with my text regardless of contract. I can only sell you usage rights - which may or may not be exclusive. If the text I wrote is trivial, neither you nor me can limit when it is reproduced. The effort of collecting data is not sufficient, if the data itself is declared trivial. See rulings about phone books.
When an AI provider produces data that is deemed not copyrightable, it cannot legally sell you exclusive usage rights. It can give it to you exclusively, but since you cannot yourself claim copyright, the moment you publish it it becomes available for others to use as well. One may argue that an LLM is similar to a phone book, with its entries being “trivial“ and its composition not artistic enough.
At least that’s the line of argument.
Ownership is not the same as authorship. Copyright is about authors rights, not owners.
If you buy a text of me, I cannot sign away my authorship, and there’s certain limitations on what you can do with my text regardless of contract.
(This is correct in many jurisdictions but not in all; for example moral rights are not a thing everywhere.)
"I paid money to someone to write that code" is exactly the line of argument that would lead to no copyright. There was the famous monkey selfie court case in 2015 that ruled that at least in the US, monkeys can't hold copyright. The same arguments would apply to AI. And since the AI can't create copyrighted works on its own, it can't assign copyright for works it created for hire either
The other line of argument is the "Claude Code is to coding like a photo camera is to painting". The image is generated automatically, but the input in how you point the camera is enough to still make it a creative work protected by copyright. Under that interpretation, you are not hiring AI, you are using it like a tool
The US Copyright Office holds the former opinion. I'm sure once this goes to court, lots of companies will vehemently argue the latter. I would not be surprised if we even end up changing the law over this
>"I paid money to someone to write that code" is exactly the line of argument that would lead to no copyright.
That's news to me. I (along with many hundreds of others) was paid to develop Minecraft, candy crush and battlefield, yet last I checked, they all retain their copyright.
Because you are human, and can thus create copyrighted works, and can assign that copyright when you do work for hire. Monkeys and AIs can't create copyrighted works, so there is nothing to assign when doing it for hire
The other line of argument avoids that issue by arguing that you personally created the code with the help of a tool (like a compiler or camera), not just commissioned it
Sure, but that's not what GP stated at all, unless possibly if they're anthropomorphizing to the point that they refer to Claude as "someone".
GP's (or rather GGGP's) statement was "I did not write that code but I paid money". So they claim no authorship. Anthropic is a company, they can't be the author either. So the only one left as author in that reasoning is Claude.
I don't think that necessarily anthropomorphizes it. We speak of monkeys as authors without calling them human. And really the legally important fact is that there was no human author. You can also treat it like CCTV footage which is generally not under copyright because there is no human author (even though most would hesitate to call the camera the author either)
As a human, it is possible for you to create copyrightable works and transfer the copyright to Microsoft in exchange for money. It is not possible for Claude Code to do those things because Claude Code is not a human.
Yes, but Claude code is hardly a "someone", so that's not what the comment I replied to was arguing at all.
They retain ownership over that code because you signed a contract saying explicitly that. Did you sign the same thing with Anthropic? Did your company?
Anthropic isn't a party to such contract either because they (most likely, in the reasonable reading of relevant laws) hold no copyright over the output of their LLM.
In this case, it's more like you paid money to FablePool. FablePool used Claude as a tool, and it delivers the product so they are the owners of the code, and have the MIT copyright assigned to them.
Nope. Copyright simply doesn't work like that. Unless there's real human creativity involved in the process, they can't just claim copyright to LLM output.
That's a problem more and more products in software will face.
In a few years most saas will have 95 percent or even more AI coded code.
Could I steal it and put it on git?
That'll translate across copyright jurisdictions.
I don’t know, if the design itself is copyrighted you could argue that the AI is just a bunch of hired workers that built it for extremely low wages.
If I hired a bunch of people to build me a house, and I drafted the architectural plans with the help of a paid architect, neither the architect nor the builders have ownership over the home.
So if a collection of people design something together maybe that has merit, they collectively paid for Anthropic to build it for them…
I’m pretty sure copyright office has settled that already. Inly human expression can be copyrighted:
> As described above, in many circumstances these outputs will be copyrightable in whole or in part—where AI is used as a tool, and where a human has been able to determine the expressive elements they contain. Prompts alone, however, at this stage are unlikely to satisfy those requirements.
https://www.copyright.gov/ai/Copyright-and-Artificial-Intell...
The United States Copyright office. There's a whole world outside the US.
And even then they can change their mind.
Does not hurt to backstop with an explicit license.
> you could argue that the AI is just a bunch of hired workers that built it for extremely low wages.
I am not able rightly to apprehend the kind of confusion of ideas that could provoke such an assertion.
With apologies to Mr. Charles Babbage.
I think pooling/donating tokens will be a thing. Not sure if like this, but in some format. The Django project, for example, came out and said they don't want your tokens, but I think a lot of people/projects will (do?) want your tokens.
Why not just give a project money and let them decide how to spend?
I guess if you have a subscription with a token allowance that you are not going to use up this week, it’s better to let someone else use those tokens rather than throwing them away. So using the food analogy it’s more like a store giving away unsold sandwiches to the homeless at the end of the day instead of throwing them away.
Donating tokens to a software project is a bit like donating food to a hungry person.
I think it might be beneficial to use blockchain, so that the donor can audit which prompts the token-pool they donated too performed. Perhaps donating tokens can also give you votes on which prompts are entered.
It’s a lot more like giving a hungry Hindu a gift card to a specific non-veg restaurant. Maybe they’ll use and enjoy it, maybe they’re vegetarian and will be insulted; either way that restaurant benefits. Especially if the hungry person exceeds the value of the gift card.
It’s more like donating snack cakes to a hungry person.
Primeagen predicted this in his latest video. I just didn't think I'll see this today.
The good ones all seem to be pointing in the direction of Django. Which, on its own, says a lot about how likely people will care about vibe-coded anything, whether pooled or not.
This is exactly how I'm building an OS right now. I have a lot of things speced out, and for most of them also create an issue. And I have a friend that just points his claudr code at the repo and tells it to "find the next thing to work on and implement it" I then do the review, verification, etc, but a great way to used unused quota.
I've always wanted to figure out how to implement a cooperative source license. Something like, you're allowed to do what you want with it, but any derivative work requires the same license, and X% of any income goes to the cooperative?
Not sure how it'd work, but there's absolutely a niche for a privacy focused data cooperative out there.
> X% of any income
Any income from what? The code is free, right? X% of your company's total revenue? Might as well just say "companies can't use this".
Personally I like the idea of a "free as in freedom but not free as in beer" license. You have to pay for a copy of the software, but after that you're free to use and modify it as you please, and share/sell your modifications under the same license.
To turn that into a cooperative you could have a company own the code and pay developers in shares of the company for PRs or other contributioins.
yeah it should really be CC0
Should it? If it was real world infrastructure, like a bridge it'd be easier to say that it belongs to those who lead the project and those who put down the money
The nice thing about a CC0 work is that it belongs to everybody. The leaders of the project have the same rights to use and modify the software as they do with software they have exclusive copyright over. In fact copyright does not grant the rights holder any new rights they did not have, it only restricts the rights of other people.
it probably already is in the public domain under US law, this just gives it the same status across jurisdictions
Highly detailed plan with time for the backers to comment and suggest improvements*
The problem with running open source code is the security aspect, but with Mythos running point, how would you distribute revenue is the real question.
Which market is even left after since the sasspocaloypse?
Maybe the financiers of a project just need it, they need it working, not to generate revenue for them?
I love how even the "demo build" doesn't work. https://fablepool.com/projects/7
Rather, it did work at milestone 14, but then regressed at milestone 15, where it changed the link from a wikimedia image to a nonexistent file in /assets (despite still having the "Photo via Wikimedia Commons" caption).
edit: they removed it :^)
If you check "DEPLOYMENT.md," there is a lengthy list of deployment instructions for the app, and it includes creating an assets folder and putting an image of Claude Shannon in it. There are also other instructions, like "please make a favicon." So I think that bit is valid, the AI is simply farming out work to the human agent.
My question, though, is why the "Live, public build log" only showing up to milestone 3, but the artifacts go up to milestone 15? And there are different index.html pages in the artifacts list, one for milestone 14 and one for milestone 15? Are there different conceptions of "milestone" in here? What's up with that?
I have found that on long-running tasks, many of the communication (and other) invariants get dropped at seemingly arbitrary points along the way. It probably just stopped doing the log.
yeah.. deployment.md had instructions to stick a photo there, but rather than explain I just got rid. gonna work on a few examples and fund them so people can see it actually work
It seems highly suspicious that all the target cost estimates are for $150-$400, regardless if it’s a bench for pelican on bicycle, or a clone of AAA game.
It reminds me of scam eshops where everything cost $random dollars in a hope that someone will enter a credit card number.
the estimation-engine is a work in progress. defs needs some work
I wrote this to a friend in 2022:
Here's an idea: reverse kickstarter
1. people post ideas
2. good ideas go viral
3. people pledge actual money to encourage someone to step forward and build it
4. interested creators make kickstarter type videos explaining their proposal for making the thing
5A. people vote on which proposal to accept, or maybe
5B. each backer can select a project to support
---
Here steps 4 and 5 are replaced by Claude.
Cool idea!
Shouldn't we still have people in the loop for selecting/proposing the best implementation (plan)? Vibe coding an entire solution from a prompt still doesn't feel like the optimal way to write software.
At some point you have to say "Is not having it better than having it?" Where's your dude, today, who's gonna code this? If it were gone happen, it wouldve.
We are not there yet... and Fable makes me feel like it will be a while before we get there.
It also occurs to me now that the Claude version has about 3 fewer zeros at the end of the funding target.
That seems to massively lower the bar for people investing.
not with prompts but "loops" maybe.
"Solve Garbage Collection in C# for HFT · $10.00 raised of est. $200.00 target"
This can't be serious.
Broader point I am making is, what differentiates genuine ideas from the token burn? What happens when the pool exhausts but the task is not done?
From my 10 years in the .net, it seemed C# devs will pretty much do anything to avoid using the right tool for the job or solving the immediate problem at hand.
C# and Java devs are anecdotally the only kind of dev to think of themselves primarily by language.
Most other devs don’t talk about language in a driest few sentences intro.
You keep putting money into the slot and pulling the lever
But at least collectively pulling it :-)
While standing around in a circle
Real question is, how do you get press for this site after this falls off HN?
It's already solved (by humans) for Java, which can now be used for HFT. It seems like it's possible to do for C#.
Yes, this is the "LMAX Disruptor": https://lmax-exchange.github.io/disruptor/user-guide/index.h...
You can write only-stack-alloc or limited-alloc C#, and Microsoft have put a lot of work into it (Span etc); it's just a bit unidiomatic.
Mind you, the last time I had contact with HFT it was inside an FPGA context..
I think assumption of the gp is that while Fable might be impressive, even Fable would take a bit more (sarcastically meaning a lot more) than $200 of tokens to solve this quite serious problem.
The sarcastic solution is to use C# bindings to a non-GC language. Put all available memory under control of a pool allocator and enjoy the perf gain.
Similar solution worked for ASP back in 1999. ASP/VBS was terrible slow at string building and Response.Write. Build it in the fast code and then output.
lol saw that one too
"A thorough written survey of why .NET garbage collection causes latency spikes in HFT contexts"
i'm like, dude, just rewrite in Zig if you want that control back, not all of your compute goodies will come from Redmond
market decides - just like kickstarter
It indicates the level of trust people have in the platform, and the combination of the product-platform behavior. If someone with the wherewithal to solve garbage collection for C# for HFT could actually describe why GC in C# was a problem, they wouldn't be asking for $10. But for $10, for something something you're dimly aware of is a problem? I'd throw $10 at some nonsense I read on the Internet.
> What happens when the pool exhausts but the task is not done?
Have a stupider LLM aggregate similar questions.
I'm not sure why folks continue to build services with trademarked names they don't own.
If Anthropic had a problem with Clawdbot they are certainly going to take issue with FablePool.
Interesting that this doesn't seem to use blockchains. Arguably it would have been a good use case. OP, could you elaborate on the reasons for the choice (if it was a conscious one at all)?
I was surprised. I expected it to be crypto payments.
Mixed in with all these aspirational positive things is some sad person trying to get a better Microsoft Teams client.
https://fablepool.com/projects/76
I'd love to see Anthropic (or someone with mythos access) create a cybersecurity version of this. So that I could create a pool that says "find security concerns in this github repo." Then the report from mythos gets sent to the code/project maintainer and revealed to the public (that paid for it) at the 90 day mark.
For your information, a group of Mythos-approved users at Apple, Google, Microsoft, and several other Project Glasswing partners have already been doing this for the past few months. We just can’t share many details publicly yet.
Who foots the bill?
The target codebase cannot improve beyond the point that the reports are incorrect and a waste of money.
There is also the question of whether humans can waste so much time reviewing AI code that the vulnerability is not patched before it is exploited. Another one is whether when the human is removed from the loop that the codebase becomes more vulnerable in some other ways.
sounds like FableBugBounty
Would love to see non-technical audiences think like this.
For years I've been trying to get estate agents to support an open-source real estate website builder. The pitch is obvious: instead of each agent paying thousands for a bespoke site, pool resources, fund the features you all need, and everyone benefits.
Getting non-technical people to commit to something abstract before it exists is nearly impossible though. Hope a model like FablePool can change that.
For the website builder, the open-source product is already there: https://github.com/etewiah/property_web_builder. It just needs momentum.
So the completed sample was estimated at $0.35, actually cost $0.52, but spend $0.55
This bot is almost as bad as I am at estimating projects.
> Build a completely greenroom, open source AWS
> est. total target $516.00
Lol
Is green room a word? I've heard clean room. And green field. Is it just an amalgamation?
It is. But doesn’t fit with the rest of the sentence.
“In a television studio, theatre or concert hall, the room where performers await their entrance.” https://en.wiktionary.org/wiki/green_room
Not sure but a green screen version of AWS for $500 is totally plausible and a decent metaphor for what you’d get with AI anyway.
Maybe a green field clean room implementation :)
Its where the band can hang out before the show.
author mistook the word
A lot of AWS is built on open-source. This is obviously ignoring hardware costs. I don’t know if it is all that ridiculous anymore. These models are very good at wiring together open-source systems. The world is crazy right now…
AWS has over 200 services, so that's a little over $2 per service. Yeah, a lot of it is built on OSS, but there is a ton of it, and there is also a lot of work involved in building the APIs and web UI, and making it scalable , secure, and resilient.
Now, you might be able to make a version of some small subset of aws services that runs works ok for a small scale for with relatively simple needs, for that many tokens, but I don't think that's what they were going for.
IAM for example is in house and integrates with every service. Sometimes in deep ways.
I’m on your side in that I would never take a contract to actually do this, but…
If we swapped out the IAM backend for something extremely simple like just private keys (one per allowed service or JWT-style list all services in the key), then we could have something that looks/feels pretty similar. With a 2$ token spend.
Not at all the same but it would look/feel pretty close.
> you might be able to make a version of some small subset of aws services that runs works ok
lol, you’ve got that goblet of koolaid with me! Equal parts horrifying and interesting that it might not be impossible
The hard parts are not based on OSS
Exactly. It's like someone saying, that EC2 is OpenStack. Well, yes, but actually no :)
But, like, on paper it is, and on paper is where the prompt lives.
You’re getting lost buddy it’s definitely ridiculous.
You, my friend, have drunk from the goblet of koolaid.
Careful dude, that's not a goblet of koolaid, it's a bottle of Amazon Delivery Driver Urine!!!
lol but at least in comes in a nice cup then
lolz. build aws. no mistakes.
Did it not charge anything for the estimation itself? I wonder what model they’re using for that
I feel like using Fable in the name is a mistake, who knows how long that model will be around.
You could call it aiproductsexchange.com
Bold move leaving out the dash between words a la experts-exchange lol.
thatsthejoke.jpg
expertsexchange.com was a site from the before times.
I wondered if it was intentional, but thought I would double down on it in case people missed it.
Was the dashless domain really a site (or the site) at one point?
It was dashless for many years.
As dashless and pointed as https://penisland.net/ !
always have a backup plan (:
Without the sex change in them : AIProductMarket.com, AIProductHub.com, AIProductMarketplace.com, AIToolMarket.com, AIToolsHub.com
That misses the point entirely....
xda-developers.com vs https://en.wikipedia.org/wiki/O2_Xda
I don't think using the name Fable is wrong, but I think a pool of Fables should be called a Grimm, or possibly an Aesop.
Perhaps Grimoire?
"A grimoire is a textbook of magic and sorcery. Traditionally, it contains instructions for casting spells, performing divination, creating magical objects like talismans, and summoning supernatural entities such as angels or spirits."
Seems to fit.
It's how they name classes of models, presumably this implies something about the relative quantization / size of model, not about the specific performance. E.g. Fabel 5 will be better than Opus 5, better than Sonnet 5, etc. The 5 is the version number of the particular iteration / training run at this class of model.
I think they mean: I feel like using [Sonnet/Opus/Fable] in the name [URL] is a mistake, who knows how long that model will be around
But it sounds like FableFool so it has that going for it.
Even if that product disappears, OpenAI will never Anthropic forget it.
We have entered the GoFundMe era of vibe coding.
goVibeMe
No
shit, that was the quickest $23 I went and bought a a domain name for.
Interesting way to waste $23
Some people have a drinking budget, I have a domain name budget. A couple years back, my friends threw a domain name transfer party, like a secret Santa where you got to surprise someone with a domain name you got for them.
Beat me to it by minutes.
You don't list contact info on your profile page to contact you with, but I'd love someone (other than Fable/a bot) to work on this with, and you're clearly a like minded individual, so reach out
I sent you an email.
Good call.
Before putting in money to this small anonymous website, I'd love to hear about the people behind the project. There's a single mention of 'Barras Industries', but not much mention about them online, or what else they've worked on.
fair comment! i'll add a link but we're at barrasindustries.com
They should have called this "WishingWell". I'm wishing them well, but some of these projects are so over the top pie-in-the-sky silly, and funded with $0.25.
Sounds like great idea, but can’t seem to find some key info: which “public ledger” is used here? A blockchain? If so, there would need to be a massive fee overhead. If not, then how are we not relying on trust?
Also, is there some kind of ownership structure based on investment?
Almost all of the examples on the website are ridiculous, which in turn makes your project look bad.
Imho you should wipe them, populate it with some realistic small scale ideas and be much more strict in review, at least for now.
The first one I saw was local first memory for AI, which seemed entirely reasonable.
I think the bottleneck is testing. I want to build a replacement for Zwift, a virtual gym game for bike trainers and treadmills, but testing it could be difficult without a real person on real hardware. How does the LLM know about the hardware protocols and stuff like that.
Same way you’d do it without AI. Record sample data, test against that, generate more data, test IRL, record more data, loop until it’s good enough.
I don't have one. how through is this blog post reverse engineering it? https://www.makinolo.com/blog/2024/07/26/zwift-ride-protocol... ?
This is genius! I can already see improved versions of this idea making it big.
Same, there is massive potential here for groups in the public guiding agents and having skin in the game with their own money.
Why are there no reverse engineering projects listed? These would make the most sense actually. Or future features of MacOS, Apple hardware designs (as Apple would not be able to patent them then)
Seems similar to open source bounties, which have been tried in the past and never succeeded.
We've seen something like 20+ years of different attempts of voluntary donations to fund open source, and it never worked. Companies barely fund anything voluntarily.
I'm taking the opposite approach with Supported Source (https://supso.org/) which is this: actually force companies to pay to use the project. Sell commercial licenses. Make it mandatory to using your software commercially. This approach works much, much better than voluntary donations.
I think there's a categorical difference between paying for long term maintenance voluntarily vs paying for something to exist. The latter works much better as the value prop is clear and you can scratch an itch. Kickstarter is similar.
Sooo.. what projects are most highly requested?
Fantastic idea for a rug pull
Yeah, already several thousands in it
> Rust Rewritten PostgreSQL
https://fablepool.com/projects/53
there's someone already doing this: https://github.com/malisper/pgrust
This is such a great idea. I often have things that im sure i dont have money for but maybe others would support it.
How about letting humans participate, not just LLMs.
> Make Fable 6
$1.00 raised of est. $205.00 target
Humans shouldn't provide estimates.
it's remarkable how easy it is to identify websites built with the "frontend-design" skill in Claude
I managed to write one that at least didnt had the font and colors (using 4.5)
Yesterday, I prompted Fable to improve the frontend to make it look different from Claude style, gave detailed examples etc. 15 minutes and $32 dollars (!) later (used cursor lol) it gave me the shittiest more claudiest website ever, basically ignoring everything I asked
Excellent idea, I see a few issues though.
First, your server is struggling. It took about 20+ seconds to respond just now, FYI.
Second, it's not obvious to me that I can get my money back if something doesn't pan out / get approved by a certain date from the homepage alone. That might make people hesitant to put anything in if they think it might get locked in there forever if the site dies / you take it down / etc.
[flagged]
This was a pre-LLM YC startup AssemblyMade which was basically this
I don't understand how that would not be a complete joke even if tokens were 2 orders of magnitude more expensive than they are.
Not affordable, unless the devs are in somewhere like Vietnam. And there's still no way they can build as fast. And still, at that price point, quality would be highly questionable. So yh this doesn't survive beyond the joke stage.
The mention of quality puts it firmly into the joke territory, indeed.
If you put that behind an API, you could sell the service much like the AI providers
And then get sued for fraud and go under, like Builder.ai
What if, and I know this is utterly batshit insane to suggest, but what if we don't lie about what we're doing?
Thats called Kickstarter
Sort of but in reverse.
If users posted ideas, voted on them and then other people built them then that would be the same. But kickstarter is the producer posting an idea for presale
I wonder why that didn’t happen on kick starter. Product hunt was kind of this. It’s actually interesting. Why didn’t this ever happen?
I think because you don't know /which/ developer you're going to get.
One interesting aspect of LLMs is that each one, weights frozen, can be thought of as a single developer whose work you have already evaluated.
The cost of finding, evaluating, and negotiating with a new human is tremenous.
It can work for students as a grant
how expensive do you think tokens are, and/or how cheap do you think a developer is?
Someone posted 20k/month Fable budgets only a few days ago. That's nearly 250k/year, which is what Oxide pays their employees.
https://news.ycombinator.com/item?id=48471771
Fable will actually finish the job.
Man, I really hope this kind of effort could be put into auditing the security situation of open source projects (via Mythos or not.)
This is literally an idea by the primegean on his YouTube under predictions. Self prophecy really with his reach but credit where it's due?
He's been right about other things before, such as this: https://youtu.be/m-bT5v5Tm7w
Slop cannons lol.
I forgot to fully describe the prompt since I already described it a bit on the title of the submission, which might be a problem. I hope the title of the submission itself is included alongside the prompt when giving instructions to the AI.
it's like a pre-vc funding round for projects
This is a genius idea, I love it!!
thank you!
This is actually kinda exciting. I threw in an open-source idea I've been playing with, and paid $25. I hope it comes back up soon or I'm going to have to put Fable on building a replacement.
This will be an excellent demonstration of what AI is incapable of.
I agree, but I do find it interesting to see what people want built and wish for.
Can built project really have MIT license? Considering MIT license still holds copyright but AI generated code cannot by copyrighted?
can I put open math problems on this site too with bounties
"Make Fable 6
The end of GitHub as we know it is near?
Why do open source collaboration? Why not a single product developer getting crowd paid to add features, solve bugs, using AI. So many businesses will see their moat wiped out.
On the macro level capitalism is winner-takes-all and Musk is the only one seriously playing the game. End game: own everything, including payments, and governments come begging and will protect him from citizen revolt. Supervillain/overlord territory.
Brilliant idea! We need consensus protocols for voting on phases. Similar to the "twitch" plays Pokemon phenomenom.
anyone who donates gets to vote (?)
Everything turns into a computer game and entertainment.
Maybe add a "Build a worm that shuts down all Anthropic data centers."
This, unfortunately, gets flagged for cyber and you would need to be on the unlocked Mythos.
I got an idea similar to this where the user can donate their tokens instead of dollars.
This is precisely what I thought the other day. TBH my idea is slightly better.
But I stopped after asking Claude about it. It categorically told me that the moment you fund a model, you are legally liable for its actions.
How to get around it?
I doubt that's legally decided yet. There hasn't been precedent as it hasn't been challenged yet.
I mean Claude will tell you because anthropic made it tell you that, doesn't mean it's true.
GoFundMe and indigogo aren't responsible for the actions of the funded projects either, hence it's unlikely that any judge would decide that the liability would go to the platform if it can show it's doing it's best effort in moderation wrt illegal content
If you mean just throw it together and then don't moderate at all then .. yeah, you'll be held liable. But that's not because of the person paying the prompt, it's because moderating illegal content is the responsibility of the platform provider.
I'd fund "clone fablepool" for $5. Should be plenty.
Really fun idea that is simultaneously deeply embarrassing for Anthropic.
Kinda fun but the approach today is strictly oneshot. Waiting for agentswithwallets to post.
same but different than this https://github.com/adv0r/tokens-for-good
This strikes me as crowd-funded prompt caching, but with humans in the loop.
Could anyone post a project to turn that site into phub for LLMs?
Hell yeah, $516 for a complete AWS replacement, I'm in lol!
Reminds of the four college kids that were going to clone Facebook. Turns out it's hard than it looks, if you have never tried it.
The coding isn't the hard part. It's the people and networking. Facebook's only moat is HOA boards that think private communication behind Facebook groups somehow equates to public messaging a community...
In other words, once people got on it, it was too late.
If you look at the milestones it's a small subset of AWS features, but yeah, the estimate is still off.
I wonder how the estimates are being created.
I doubt an LLM would estimate an AWS rewrite to cost $500.
attach github to this. this is the new way to do opensource i guess
that's the plan! funded projects will spin up a repo
"Build a completely greenroom, open source AWS" – $700
This is engineering theatre (pun intended).
The amount of hubris here is exceptional, the author doesn't even know that it's "clean room" rather than "green room". What does it even mean to build an open source AWS? There are many open source IaaS/PaaS components. Is the author suggesting any hardware design, because that's a critical component.
The only possible result of this is an AWS fanfic. An art project that looks vaguely like a cloud provider on the surface if you squint, but with zero substance to it.
And this criticism has nothing to do with AI. You'd get the same spending 100x that budget on any engineering team.
Has anything been successfully built?
Fable’s been out for like a day and this site seems more recent.
fablepool didn't exist 24 hours ago .. so not yet
Do we need a cryptocurrency for trading / donating LLM compute tokens ?
Made something very close to this, but not model specific. Ill try to shape it up tonight and tmr and drop it, would be cool to colab!
full disclosure, it's model specific because the domain was available + bandwagon
Like DeFi but for agencies.
Is this the new open source?
At least it recognizes that energy deserves funding, ideally beforehand. Yet it would be harder to sell if a human asked for payment, even if delivery was guaranteed.
This is DOA. Claude will refuse to work on anything there or better just bankrupt the funding members while delivering slop(on purpose to doge their cyber operations )
https://fablepool.com/projects/7 It didn't even put a picture in!
"Build Grand Theft Auto 7" I like that here are my 0.25c
A heads up, since LLM is not itself responsible for any wrong it does, its operators are, and you are that guy.
Oh
I could see something like this working if you actually had a assigned human developer(s) to assist the task. There are few interesting tasks that can actually be completed in one (or few) shot and have anything usable.
I need an X11/Wayland successor that has the simplicity of X11 but can be used assl a drop in replacement for Wayland
Isn't wayland a protocol? A drop in would carry the complexity.
Awesome idea.
thank you!
It seems weird that you would have about 3% of your revenue taken away by card providers you should just accept USDC.
Have any successful funds?
It would work if an engineer steered the pools. But doing this autonomously is a pipe dream.
there's a job going if you fancy it... pool steersman
We have a saying in Australia: "Up shit creek without a paddle."
Cypherpunks will be proud once there is a version of this cryptocurrency funded to providers receiving the cryptocurrency.
Or maybe there is? or a version where only those funding have access to the results.
This is a good idea and for features and modifications you can make it so whoever chips in the most money gets more votes.
This is one of those ideas that sounds bad on paper (Like people renting out their houses. But if implemented correctly could get some traction.
thank you!
This is a fantastic idea.
There are lots of projects, software that shouldn't be SaaS subscriptions that Fable can build in public that can be free for everyone and also OSS.
Neat project idea, but truly ruined by requiring a google sign-in both to submit new projects and to donate to projects. Dead service to me until that's gone.
Remember, Google aids and abets militaries of governments that the UN has found to be committing genocide.
Weird how people seem to forget this.
> Dead service to me until that's gone.
Lets just hope the project is able to soldier on without you.
Google was just the easiest to implement first. Was planning Github next - or would you prefer smth else?
My first choice would be non-SSO. Let me pick a username (or email address) and a password.
Btw, thanks for the response. Not sure why you got downvoted for it, but you have my gratitude for being one of few devs who are sincerely responsive to these types of concerns.
Lol good place for multiple eyes to view how limited "ai" is.
Ok who wants to pool up to build GTA 7? /s
This is such a good idea. Hell yeah
Lol.
"I want an open source AWS" with $500 budget made me guffaw
"I have a turbofan model, pls build an Airbus" sounds about right
All for $670 :)
In all seriousness, I would probably throw $10 at a project to design and implement a modern turbofan FADEC + all of the certification artifacts.
I got downvoted probably for tone which is my bad but what I’m laughing at is this person doesnt seem to understand why people pay for aws, it’s certainly not the laughably bad console or the buggy control plane. it’s the reliability guarantees granted by their massive physical infrastructure that was meant to replace sysadmin’s running racks in a closet and wrangling terrible ansible/chef playbooks.
this literally already exists if you’re willing to maintain your own physical infra, and has for a long time - nothing aws does is that innovative software wise. maybe their managed k8s eliminates a ton of pain, but I dont know. it’s the reliability guarantee + support + not having to maintain physical servers. if youre willing to shirk all that and do it yourself why would you want aws? lol
tldr; was laughing at the xy vibe of the ask
OpenStack already exists