Here's a broken rule almost everyone accepts without thinking: to get funded, you pitch. You describe a thing that does not exist, you promise it will work, and someone decides whether to bet on your promise. The money flows to the best story. Whether the thing ever gets built, and whether it's any good, is a later problem — often someone else's.
Now flip it. What if you paid for the good after it existed, once everyone could see it worked and see who it helped? No pitch. No promise. Just: this was built, it's valuable, here's your reward.
That's retroactive funding. And once you see it, the old way starts to look insane.
The core inversion
Traditional grants and venture money reward predicted value. You are betting on the future, which means you are betting on stories, credentials, and charisma. This is a bug the powerful have always exploited — the best-connected pitch, not the best-outcome, wins the capital.
Retroactive funding rewards realized value. The judging happens after the fact, on evidence. Did this open-source tool get used by thousands? Did this piece of public infrastructure hold up? Did this research actually get cited and built upon? You measure what happened, then you pay for it.
The insight underneath is deceptively simple, from the crypto world that coined the phrase: it is much easier to agree on what was useful than to predict what will be. Hindsight is a cheaper oracle than foresight. Arguing about which unbuilt project deserves money is endless. Looking at ten shipped things and ranking which mattered most — that, humans can actually do.
Where this already runs
This is not theory. Optimism, an Ethereum layer-2, has run several rounds of what it calls Retroactive Public Goods Funding (RetroPGF / Retro Funding), distributing tens of millions of dollars' worth of tokens to people who had already built tools, infrastructure, documentation, and education that the ecosystem depended on. Nobody pitched a roadmap. Badge-holders looked at what existed, at what got used, and rewarded it.
The mechanism is roughly:
- A pool of funds is set aside for a round.
- Anyone who built something useful in a defined scope can be nominated.
- A set of voters — reputation-weighted, chosen for judgment, not for stake size — allocates the pool across the contributions by observed impact.
- Payment flows to the builders, after the fact, for work already done and already validated by reality.
Public goods — the things everyone benefits from and no one individually profits enough to build — have been chronically underfunded for all of human history. This is the first mechanism at scale that pays for them because they turned out to matter, not because someone gambled they might.
Our Record
Weigh the deed, not the boast. That is the whole logic of the Scales of Ma'at — the heart is measured against the feather after a life is lived, on what was actually done, not on what was promised at the start. Isfet is the system that pays the loudest pitch and lets the quiet builder starve — value extracted from a story, not from a good. Retroactive funding drags the reward back to the deed itself: build the thing, let it prove its Ma'at in the open, then be paid for the weight it actually carried. It rewards contribution over speculation, substance over performance. The feather does not care how well you described the good. It cares whether the good is real.
The honest limits — because there's no free lunch
Retroactive funding is powerful, not magic. Be clear-eyed:
- Builders still need to eat while building. If reward only comes after impact is proven, who funds the months in between? The honest answer: retro funding pairs best with something that carries you through the build — savings, salary, upfront grants, or streaming income. It rewards outcomes; it does not, by itself, solve cash flow. Mature designs mix retro with some up-front support.
- Judging impact is genuinely hard. "Useful" is easier than "will be useful," but it's not objective. Voters have blind spots and biases. Flashy, visible work can beat quiet, foundational work — the same trap as before, just later. Good rounds fight this with diverse, thoughtful judges; bad rounds just recreate a popularity contest.
- It can be gamed. Where there's a pool, there's someone farming it — fake impact, sybil nominations, manufactured metrics. Sybil resistance and human judgment are the defense, and both cost effort.
- It doesn't fund the truly unproven moonshot. Some things need a big bet before anyone can see if they work. Retro funding is weak exactly there. It's a complement to forward funding, not a full replacement.
- The round's scope decides everything. Whoever draws the boundary of what counts as "useful" has already baked in politics. A narrow scope rewards insiders; a broad one dilutes judgment. Whoever holds the pen that draws the circle holds part of the power. Not an argument against the method — a reminder to watch who's drawing the line.
None of that sinks the idea. It just means retro funding is a tool with a shape — use it where realized value can be seen and weighed, pair it with something that carries builders through the work.
The lever
If you run a treasury, a grant program, a DAO, or you're deciding how to reward the people around you — try inverting one funding decision. Instead of asking "who has the best plan," set aside a pool and ask "what already got built here that mattered, and who built it?" Pay for the deed after you can see its weight.
If you're a builder tired of writing pitch decks for work you'd rather just do — go do it, ship it in the open, let it accrue evidence, and seek the ecosystems that reward what's real. They're growing.
The old system paid the best storyteller. The new one can pay the best builder. That's not a small fix. That's dragging the reward back onto the deed — where Ma'at always said it belonged.