Owocki is a fan of quadratic funding, but it has big deficiencies:
Many people do not know that they need Hilbert spaces theory or molecular biology very much, word processor users don’t know that they need a reliable optimizing compiler that is in turn may be based on dependent-types lambda-calculus research, etc. and they won’t fund it. Especially science and software libraries public goods suffer severe underfunding when funding quadratic, because every funder cannot know every science topic.
When doing quadratic funding it inevitably goes much random - some projects are funded too much and some are severely underfunded (among other reasons, because well-funded projects can buy SEO of their ads).
Anti-sybil is improving but it can’t be perfect.
If some important text or software is buried on the 100000th position of Google, the world gets stuck: nobody knows that the missing component is a problem and therefore nobody fixes the problem.
So, we need a much more objective mean of funding. Ideally funding amount should be calculated by an algorithm, not by stupid us.
We don’t yet have an algorithm capable to predict the future of the human work. The solution is to use a prediction market to predict a future algorithm results. So the market decides.
Traders (when motivated by money) are much more careful, hard working, and through than some random quadratic voter. Traders are therefore multitudinous that much reduces randomness and bias. They are to follow an objective algorithm that motivates them and so the market is striving for objective results.
So, I provide a software that “forces” traders to “calculate” how much which project to fund.
There will appear a new profession: people that search for underrepresented public good projects and do SEO for them. That’s important.
It should be in addition to quadratic Grants. Users should have choice where to pay.
This software is almost fully implemented (and there is a little outdated pull request to integrate it into a GitCoin page).
The proposal:
Here is the PR and more details: