TL;DR:

We propose a set of “Legos” for effective, strategic funding of common goals, based on first principles from collective intelligence theory and supported by real-life experience from large public and private organizations.

<aside> ➰ Seven Legos for funding common goals

1. End goals are quadratically prioritized.
2. Collaborative and evolving goal graph.
3. Projects contribute to goals.
4. Funding backpropagates along goal graph.
5. Retroactive funding through collaborative impact assessments.
6. Individuals contribute to projects and accrue cred.
7. Systematic review and adaptation of the overall design. </aside>

We believe that implementing funding mechanisms based on these Legos can allow the world to scale common goal funding to the level of $trillions per year allocated within the next decade.

Background

The authors are entrepreneurs, collective intelligence researchers and practitioners with combined XX years of experience in strategic goal-setting and resource management in public and private organizations ranging from 3 to 300,000 people…

The problem framing for organizations isn’t exactly like the one we’re focusing here, but it’s not that different either: dilemmas like balancing global priorities and crucial feedback from the edges, resisting centralization and popularity contests in decision-making, and sharing state across a large and diverse network of people with competing priorities - these are all active challenges for organizations as well, and have established patterns that we can learn from.

Problem

There is a substantial amount of resources out there that’s intended to improve humanity’s welfare. These are distributed into pools, from government budgets to foundation grants and philanthropic endowments. The question naturally arises of how to best allocate these resources so that they actually achieve the end goal. The **“pure-quadratic design paradigm”** for common goods funding has gained attention since its 2018 proposal by Vitalik Buterin, Zoe Hitzig and Glen Weyl, and has been implemented to allocate several $M, most notably through Gitcoin.

Here we propose a somewhat different paradigm, where the purpose is to fund common goals. This alternate framing highlights various additional constraints, that are abstracted away by the “standard model” justifying the pure-quadratic paradigm:

• Typically, an individual cannot make reliable assessments of an arbitrary proposed public good’s value to themselves. They may know what they want generically (ex: less air or noise pollution, better hospital service, more green spaces with nice benches etc), but they cannot even come close to parsing the vast amount of possible public good baskets that might accomplish them, even if they had good information about each possible item in that basket, which they usually don’t.
• Many or most public goods look like projects, not like consumer goods; i.e., the project is either successful and accomplishes value for a large swath of people (including indirect value by enabling several other goods), or it doesn’t. Thus, this is actually a portfolio problem involving non-trivial dependencies and synergies.
• Funding is a necessary but not sufficient condition for project success. In addition to upstream dependencies (mentioned above), having the right team, well-posed project framing, an appropriate process and all manner of exogenous conditions also factor in. By the way, projects are often lumpy: funding 50% of a project doesn’t provide 50% of the value, but 0%. At the same time, project insiders have incentives to overstate their needs.
• If projects are proposed bottom-up rather than determined top-down by following from individuals’ desires, there are additional issues of matching projects to goals, self-selection and deduplication.
• Resource supply (either individuals’ contributions or the matching pool) is not a given, but rather a function of the expected value generated by the portfolio. These expectations can be inflated to attract more resources, but this incurs a risk of disappointment and cynicism. A common result is the “curse of good operational discipline”, where project ambitions are systematically curtailed to what they can deliver with “100% confidence” and project plans are based not on what’s needed to succeed but what can be justified through the accounting process.
• Finally, there are all sorts of incentive problems associated to how individuals are rewarded within project teams. These problems can feed back into the project selection mechanism and, in extreme cases (such as many countries’ public budget processes), turn the portfolio into a pure machine to divert resources to project insiders, with no net value created.

These constraints are well-trodden ground in disciplines from institutional economics to organizational theory, finance, decision science, etc. They add substantial requirements to the solution space. Nonetheless, the implicit perspective in the pure-quadratic paradigm is that we can start from a minimal design justified by a stylized problem framing that ignores them, and adding features to tackle these and other constraints; the 2018 paper already tacks some such variations on, and practical implementations like Gitcoin’s already show the need for adding even more. The results are brittle designs that are hard to reason about, hard to iterate on, and hard to scale. Instead, we claim that a holistic design perspective can provide a better starting point by naturally and harmonically addressing these constraints.

The perspective we propose follows directly from interpreting the problem as an actual (collective) intelligence iteratively learning its own goals and how to achieve them. We can use that as a scaffolding to interpret various design elements as canonical implementations of fundamental capabilities of an agent in a complex, multipartite internal and internal environment.

Vision

In 10 years, $1T/year will be allocated to funding common goals, through composable, reliable processes that enable scale while preserving outcome integrity.