The Core Idea
Universal Basic Income answers the right question—how do people survive when AI displaces work?—but it may answer it too bluntly. UBI says: everyone gets the same amount, no questions asked.
Proof of Benefit asks a different question: what if we could measure and reward the value people create, even when that value doesn't fit neatly into a traditional job?
The insight is that "work" has always been broader than "employment." Parenting, mentoring, caregiving, community organizing, open-source software, local journalism, civic participation—all of these generate enormous value. The economy just has no mechanism to compensate them. In a world where AI handles more of the tasks employers currently pay for, these human contributions become more important, not less.
“Proof of benefit is basically pricing. If you can demonstrate the value, you can justify the reward.”
The Blockchain Analogy
The name isn't accidental. In cryptocurrency:
- Proof of Work rewards computational effort—whoever burns the most electricity wins.
- Proof of Stake rewards commitment—whoever has the most invested wins.
Proof of Benefit would reward demonstrable positive impact—whoever creates the most value for others receives the most support.
This doesn't require blockchain technology (though it could use it). The analogy is about the mechanism design: creating a system where the incentives align contribution with compensation, even when there's no employer writing a check.
How It Differs from UBI
| UBI | Proof of Benefit | |
|---|---|---|
| Who receives? | Everyone, unconditionally | Everyone who contributes (broadly defined) |
| Amount | Flat, same for all | Variable, based on demonstrated value |
| What counts as "work"? | N/A—no work required | Expanded far beyond employment: caregiving, mentoring, civic participation, creative work, community building |
| Incentive structure | None (that's the point) | Do more of what helps others → receive more |
| Philosophical root | Rights-based ("you exist, therefore you deserve security") | Contribution-based ("society should recognize what you give") |
| Political appeal | Left: dignity. Right: simplicity. | Left: values unpaid work. Right: maintains link between effort and reward. |
Precedents & Precursors
This isn't an idea from nowhere. Several existing models explore the same territory:
Participation Income (Anthony Atkinson, 1996)
The late Oxford economist proposed a "participation income": like UBI, but requiring recipients to make a broadly defined social contribution—employment, education, caregiving, volunteering, or active job search. Atkinson saw it as a politically viable bridge between unconditional UBI and traditional welfare's work requirements.
Status: Theoretical, never fully implemented
Open Source & the Reputation Economy
Open-source software development is a functioning proof-of-benefit system. Linus Torvalds doesn't get paid per line of code by a corporation. He (and thousands of contributors) build Linux because it's useful, and reputation, career advancement, and corporate sponsorship flow toward demonstrated value. GitHub's contribution graph is literally a proof-of-benefit dashboard.
Status: Active, but limited to tech sector
The Artisanal Software Model
An emerging pattern: individual developers build tools for their community (friends, local organizations, niche audiences) rather than scaling to millions of users. They're rewarded through a mix of patronage (Patreon, GitHub Sponsors), reputation, barter, and goodwill. It's the pre-industrial guild model, digitized.
Status: Emerging, growing rapidly with AI-assisted development
Time Banking
Community systems where people exchange hours of service: I teach your kid piano for an hour, you fix my plumbing for an hour. One hour = one hour, regardless of the market value of the skill. Over 500 time banks operate in the U.S. today, typically in local communities.
Status: Active, small scale
Social Impact Bonds
Governments pay private organizations only when they achieve measurable social outcomes (reduced recidivism, improved graduation rates). The mechanism is pure proof-of-benefit: demonstrate the result, receive the funding.
Status: Active in dozens of countries since 2010
The Hard Part: Measurement
The obvious objection: who decides what counts as "benefit"? This is the central challenge, and it's worth being honest about how difficult it is.
What's Easy to Measure
- Education: Tutoring hours, students mentored, courses completed, certifications earned
- Caregiving: Hours of eldercare, childcare, disability support provided
- Civic participation: Volunteer hours, community board service, election work
- Creative output: Works published, performances given, open-source contributions
- Environmental impact: Carbon offset, habitat restoration, cleanup events
What's Hard to Measure
- Quality vs. quantity: 100 hours of bad tutoring isn't worth 10 hours of great mentorship
- Intangible contributions: Being a good neighbor, maintaining social cohesion, emotional labor
- Gaming: Any measurement system creates incentives to optimize for the metric rather than the outcome (Goodhart's Law)
- Power dynamics: Who sets the criteria? A government agency? An algorithm? Community vote?
This is exactly where UBI advocates push back: the entire point of "unconditional" is to avoid the bureaucratic overhead and moral hazard of deciding who deserves what. Proof of Benefit trades the simplicity of UBI for the precision of targeted reward—and that trade-off has real costs.
Where AI Might Help
Ironically, AI—the same technology creating the problem—might help solve the measurement challenge:
- Automated impact assessment: AI could evaluate the downstream effects of contributions (did that tutoring actually improve test scores?) rather than just counting hours.
- Fraud detection: Pattern recognition to identify gaming without requiring a human bureaucracy.
- Peer validation: AI-mediated reputation systems where community members confirm contributions, similar to peer review in science.
- Dynamic pricing of social goods: An AI system could estimate the market-equivalent value of caregiving, mentorship, and community work in real time.
The Hybrid: UBI Floor + Proof of Benefit Bonus
The most pragmatic version may not be either/or. Consider a layered system:
Layer 1 — Universal Floor: A modest UBI (at or near poverty level) ensures no one starves. No conditions, no questions. This is the safety net.
Layer 2 — Proof of Benefit Bonus: Additional income available to anyone who demonstrates social contribution. Broadly defined, AI-assisted validation, variable payout. This is the incentive layer.
This hybrid gets you:
- The security of UBI (no one falls through the floor)
- The incentive alignment of proof-of-benefit (contribution is recognized and rewarded)
- The political viability of maintaining a connection between effort and reward (the concern that kills pure UBI for most conservatives and many moderates)
- An expanded definition of work that includes the things humans will still do better than AI: care, create, connect, teach
The Artisanal Economy
One compelling vision of how proof-of-benefit could work in practice is the artisanal model: instead of everyone working for large employers, people create tools, content, and services for their immediate communities.
What This Looks Like
- A retired teacher builds an AI-powered tutoring system for her neighborhood and is recognized (and compensated) for student outcomes.
- A developer creates a scheduling app for the local Little League and is rewarded through the community's proof-of-benefit pool.
- A caregiver's daily work supporting an aging parent is formally valued at market rates, funded through the system rather than absorbed as unpaid labor.
- A citizen journalist covering city council meetings receives compensation proportional to readership and civic engagement impact.
Why AI Makes This Possible
Before AI, building a tutoring platform or a scheduling app required specialized skills and significant time investment. AI collapses the skill barrier: anyone with a clear idea of what their community needs can build it. The bottleneck shifts from can you code? to do you understand the problem?—and local knowledge becomes the scarce, valuable resource.
Open Questions
This framework is a starting point, not a finished policy. The hard questions remain:
- Who administers the system? Government? A decentralized protocol? Community-run DAOs? Each option has different trust and corruption profiles.
- How do you prevent a social credit system? There's a fine line between "rewarding contribution" and "scoring citizens." The system must reward positive action without punishing inaction.
- What about people who genuinely can't contribute? Disability, severe illness, age extremes. The UBI floor handles this, but the framing matters: these people aren't "failing" the system.
- Can measurement scale? Proof-of-benefit might work beautifully in a community of 500 where everyone knows each other. Can it work for 330 million people?
- Does variable reward just recreate inequality? If some contributions are valued more than others, you're back to income stratification—just with different winners and losers.
Why This Matters for the AI Transition
The deepest version of this argument: if AI truly delivers abundance—cheap goods, cheap services, cheap energy—then the scarcity that remains is meaning. People need to feel useful. They need to know their contribution matters.
UBI solves the material problem. Proof of Benefit attempts to solve the existential one: in a world where machines can do most things better, faster, and cheaper, what makes a human contribution worth recognizing?
The answer might be: the things only humans can do. Care. Judgment. Creativity. Local knowledge. Moral reasoning. Community presence. These are precisely the contributions that traditional employment has always undervalued—and that a proof-of-benefit system could finally make visible.
“The most robust protection against poverty comes from institutions that generate a harmony of interests rather than those that foment distributional conflicts.” — Peter Boettke & Adam Martin