A friend wrote me a few weeks back. Have you heard of x402 payments? I had not. By Sunday evening I had the bones of a platform built. By the next week I was filing a $150K grant application to fund deeper work on it. The platform is GiveReady, and I'm trying to discern whether this is a big distraction or not. So, I'm putting it down in public about how to deal with this because the work train is starting to leave the station, and I need to be ruthless with my time to make sure I don't get sucked into project drift.
Last week I alluded to the grant and the platform without naming either, in a post about the supply side of youth adventure grants. The supply side, the grantees applying, is mostly solved. The demand side, getting donations into small funds without bank friction or a middleman layer, is the harder problem. GiveReady is my bet on solving it.
The problem statement is depressing in plain numbers. 85% of all charitable donations go to the top 10 to 12% of charities by size. The little guy, the WEF, the local youth fund, the unincorporated charity in Bermuda or Senegal or Bolivia, gets the leftovers. AI agents are about to make that worse before they make it better, because the default donation flow they recommend will be whichever charity Wikipedia ranked highest. GiveReady is an attempt to put small charities on a B2A surface where AI agents can find them and route donations without a human fundraiser layer in the middle.
Saying it out loud feels strange because for years my default would have been to keep the idea quiet. Scarcity mindset. Someone will steal this if I post about it. Two things changed. One, ideas are not the asset; execution is. Two, x.com is currently producing roughly a trillion startup ideas per week and most of them are better-positioned and better-funded than mine. Anyone with the appetite to chase a more lucrative ICP than "broke underfunded micro nonprofits" will find more lucrative ICPs. I am not worried about competition for the under-resourced corner of the donation market.
If you are reading this and you DO want to take the idea sideways into a higher-margin market, please do. The pattern (agent-discoverable directory, frictionless on-chain payment, no-fundraiser pull-through) ports cleanly to other categories. I would even cheer for it. The under-broke-charity version is the version I have personal motivation to build, because it is tied to the Finn fund.
This week, Chase and I are starting the US 501(c)(3) application for the fund, at least that's what we decided. It's if I get to it or not this week, now being a bit behind on client work. Why now? US is the largest charitable-giving market in the world, and we want US donors to be able to deduct their donations the moment the donation happens, not after we send a confirmation email three weeks later. A specialist firm runs the filing for us at $2,500. Easily worth it for the credibility lift and the tax-deductibility unlock. The fund is currently an unincorporated private grant-making fund established in Bermuda; we are currently applying for Bermuda registered charity status, and we will be applying for US 501(c)(3), which is the next layer. Hopefully that happens in 2026.
The pipe dream, to put a number on it: a $2M endowment with conventional fund management, growing at 10% per annum, with an AI-automated donation engine on top of it bringing in incremental funds to self-perpetuate the youth empowerment grants. Geographically agnostic. Whatever country a grantee comes from. Whatever dream. The fund decides on the merit of the application; the donation flow runs itself.
The supply side, the grantees, is mostly solved. The remaining work there is belief. People do not yet believe a fund will hand them grant money to follow a dream with no strings attached. I think that solves itself with enough stories of past grantees: video, written, photo. Not yet a real bottleneck because we are not yet at scale; will become one if we push for application volume without the testimonial bank to back it up.
The demand side, the donors and the engine that brings them in, is the next challenge. That is what GiveReady exists for. But that's a big IF, and if the platform starts to take off.
So what is giveready.org? Three things stacked.
One, a donation payment gateway that the WEF can use immediately. As I wrote in the charity-governance post, banks in Bermuda are not great for any kind of digital donation flow. I am solving my own problem first.
Two, the long-shot grant application. I will likely not get it. If I do, GiveReady becomes a much bigger mission and most of my next twelve months tilts toward it. If I do not, I keep the platform alive on 15 to 20 minutes a day, some days more, reading the AI-agent logs and the MCP server reads, tightening the agent-discovery surface incrementally. Both paths are fine.
Three, a B2A directory and donation engine that small charities anywhere in the world can plug into. I am in conversations with a UK consortium partner who could open doors to a large group of UK charities. If the platform shows traction, the business model figures itself out (probably a small fee per agent-routed donation, or an admin layer for charities that want extra features). If it does not show traction, it stays as the WEF's own gateway and that is enough.
Crossover with TestVentures: warm intros to charities that need fractional AI ops at 50% of standard pricing. A donation engine on the TestVentures site itself, with traffic-driven voluntary donations going direct to the WEF in one click via x402 USDC on Solana. That part is already built. Maybe the first hire I was speccing on the plane back from Madrid ends up helping with this. Maybe a reader who happens to be a CTO-caliber developer ends up wanting in. Open invitation.
My favourite saying applies here. Most people overestimate what they can do in a day and underestimate what they can do in a year. With AI compressing the build cycle by roughly an order of magnitude, giveready.org could realistically end up being the long-tail project that cannibalises the entire momentum of the next twelve months. I am okay with that outcome. The point is to do something useful in Finn's name. If GiveReady gets there faster than the fractional AI ops practice does, that is a feature, not a bug.
Coming out about the project feels good. It might not make money. It might not get the grant. The bigger version might prove too ambitious for a one-person stretch with two clients and a daily BIP commitment. None of that changes why I am building it.
Monthly Revenues $11,800 | Clients 2 | Prospects 1 (will book once closed)
Day 34 of 365.