Show HN: FablePool – pool money behind a prompt, and Fable builds it in public
fablepool.com274 points by matthewbarras 6 hours ago
274 points by matthewbarras 6 hours ago
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.
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'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.
if fable is writing it, courts my declare that its not even public domain? not a copywrightable work
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.
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?
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.
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.
Highly detailed plan with time for the backers to comment and suggest improvements*
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.
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?
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
i built a turbofan
https://app.confbuild.com/p/z459
now I want to build a complete Airbus as detailed as possible with give budget"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.
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.
You keep putting money into the slot and pulling the lever
Real question is, how do you get press for this site after this falls off HN?
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.
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.
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.
Did it not charge anything for the estimation itself? I wonder what model they’re using for that
> Build a completely greenroom, open source AWS
> est. total target $516.00
Lol
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.
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
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
Without the sex change in them : AIProductMarket.com, AIProductHub.com, AIProductMarketplace.com, AIToolMarket.com, AIToolsHub.com
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
We have entered the GoFundMe era of vibe coding.
goVibeMe
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.
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.
This is genius! I can already see improved versions of this idea making it big.
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... ?
Can built project really have MIT license? Considering MIT license still holds copyright but AI generated code cannot by copyrighted?
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.
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.
This will be an excellent demonstration of what AI is incapable of.
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
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.
Fantastic idea for a rug pull
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 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.
Really fun idea that is simultaneously deeply embarrassing for Anthropic.
Man, I really hope this kind of effort could be put into auditing the security situation of open source projects (via Mythos or not.)
[flagged]
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.
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.
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.
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
"Make Fable 6
I got an idea similar to this where the user can donate their tokens instead of dollars.
Could anyone post a project to turn that site into phub for LLMs?
Brilliant idea! We need consensus protocols for voting on phases. Similar to the "twitch" plays Pokemon phenomenom.
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.
This strikes me as crowd-funded prompt caching, but with humans in the loop.
It seems weird that you would have about 3% of your revenue taken away by card providers you should just accept USDC.
attach github to this. this is the new way to do opensource i guess
"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.
Kinda fun but the approach today is strictly oneshot. Waiting for agentswithwallets to post.
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
Has anything been successfully built?
Hell yeah, $516 for a complete AWS replacement, I'm in lol!
If you look at the milestones it's a small subset of AWS features, but yeah, the estimate is still off.
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.
I wonder how the estimates are being created.
I doubt an LLM would estimate an AWS rewrite to cost $500.
Like DeFi but for agencies.
Have any successful funds?
Is this the new open source?
It would work if an engineer steered the pools. But doing this autonomously is a pipe dream.
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.
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.
Lol good place for multiple eyes to view how limited "ai" is.
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?
[flagged]
[dead]
[dead]
[dead]
[dead]
Ok who wants to pool up to build GTA 7? /s
"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