Funding What Matters in the Age of AI
TLDR
- Gitcoin’s core mission has always been funding what matters, and our “how” has been building coordination infrastructure for public goods (most notably, QF).
- We now see the AI transition as the next major public goods challenge: concentrated power, collapsing trust, cognitive dependency, and mass displacement are all coordination failures.
- Resilience itself is a public good: communities need shared infrastructure to preserve agency, trustworthy information, local livelihoods, and accountable governance.
- We believe the tools we’ve built for Ethereum public goods funding (QF, grants infrastructure, sybil resistance, community capital allocation) can now be applied more broadly to help society coordinate through the AI transition.
Gitcoin has spent eight years building coordination infrastructure to fund public goods. The tools and the context have evolved over time, but the underlying mission hasn’t: solve coordination failures in ways that distribute power; do this by funding what matters.
The question we’re sitting with now is where that coordination capacity is needed most next.
Our read: the AI transition is producing a compounding set of harms at the scale of existential threat. Building the human resilience required to mitigate this threat is fundamentally a coordination problem, and it’s one of the fastest growing problem spaces we’ve seen.
The problems Gitcoin was built to solve – fragmented funding, captured institutions, communities that need to coordinate faster – have never been confined to Ethereum alone. These are global problems, and we think the AI transition is making them urgent at a scale that demands we export what we’ve learned to the world.
Our conviction is pointing in this direction. This post shares how we’re seeing the landscape and opens the floor to the community for feedback.
As we point in this direction, we celebrate our past: the Ethereum public goods ecosystem we’ve helped build is rich, and our early legacy lives on in the new guard of teams and protocols carrying that work forward. But we do not rest on our laurels. If we are successful in this new direction, then this past will be just a footnote in Gitcoin’s journey. What the Apple II was to the Macintosh, our early QF era could be to our AI era.
Where We’re Coming From
Over $60M distributed through Gitcoin Grants alone. Thousands of builders funded. Quadratic Funding, Grants Stack, Passport, Allo Protocol. Real coordination infrastructure, shipped and battle tested across multiple cycles.
We’re proud of this legacy. And that work gave us something specific: nearly a decade of hard won knowledge about how communities coordinate, how capital flows to what matters, and what breaks when those systems fail.
That history is our responsibility to steward forward.
The AI Transition Is Mid-Flight
The AI transition is happening in real time. Coinbase laid off 700 in Q1. Cloudflare 1,100. Meta 8,000. Brian Armstrong has publicly framed the move as “rebuild Coinbase as an intelligence, with humans around the edge.” Block has modeled a 40% headcount reduction by the end of 2027. Madison Reed 30%. Coinbase another 14% on top. Big tech capex for 2026 is $725B, almost all of it flowing into compute. Very little into labor.
The public conversation has coalesced around job displacement. It’s concrete. It’s measurable. It’s accelerating. The more time we’ve spent sitting with it, the more we think displacement is the leading edge of something broader. A compounding erosion of human agency across several fronts at once.
The Four Harms to Humanity
Four patterns keep showing up in our work, in conversations with our research community, and in the data.
Skill erasure. Knowledge work hollows out as models climb the percentile curve. The most immediate harm and the most visible.
Cognitive atrophy. As more cognition gets offloaded to AI, the capacity to think and judge independently weakens. The harm that’s hardest to measure, but potentially harder to reverse.
Collapse of shared truth. Synthetic media, recommender systems that fragment epistemic ground, and AI generated content saturate every information channel. The harm that makes coordination structurally harder.
Monopoly capture. A handful of frontier AI labs and platform companies sit between every meaningful workflow and the humans doing the work. The harm that closes the exits.
These public bads compound. Displacement pushes people into deeper dependency on systems they can’t exit. Dependency erodes the cognitive capacity to recognize capture. Capture floods the information environment so coordination against it gets harder. Fewer alternatives means more displacement.
Call it a ratchet: each harm tightens the others and the cycle only turns one way.
Job Displacement Is the Leading Edge
Job displacement is where most people are encountering this first. It’s where the felt sense kicks in. The institutional responses forming around it (UBI proposals, retraining programs, reskilling funds) tend to address income loss while leaving the structural question untouched: who governs these systems, and on what terms.
A displacement crisis managed through subsistence payments from the same entities that caused the displacement pays the bills. It doesn’t return agency.
Retraining carries its own problem. AI capability is moving faster than humans can certify into new roles. The destination skill that’s safe to invest 18 months learning is a moving target. For many roles there’s no destination at all.
The frame we think is worth exploring is broader: resilience.
What We Mean by Resilience
Resilience is the capacity to maintain agency, livelihood, and coordination capability as AI reshapes the conditions under which all three operate. It maps to the four harms above.
Economic resilience. Livelihoods that don’t compete on being faster or cheaper than a model. Work that’s rooted in place, relational, locally legitimized, or owns its own capital base.
Cognitive resilience. Independent judgment. Embodied skills. The practice of thinking without an assistant in the loop. The muscles you lose if you don’t use them.
Epistemic resilience. The capacity to coordinate around shared truth when synthetic content is indistinguishable from authentic. Requires defensive tooling and trusted small scale networks simultaneously.
Institutional resilience. Governance and funding systems that can’t be captured by the entities they’re meant to regulate. Coordination infrastructure that stays accountable to communities rather than platforms.
Resilience is a public good.
This Is Accelerating. Fast.
Model capability is improving at a pace that compresses every timeline. The gap between “AI can do X poorly” and “AI can do X at 90th percentile human level” is shrinking fast. The corporate incentive structure rewards speed over safety.
The window matters. The communities, skills, and coordination capacity we’d need to respond are being eroded by the same thing we’re trying to respond to. The response gets harder while the problem gets bigger.
The Overton window on AI displacement is open right now. It closes when a new normal sets in and people stop seeing displacement as worth organizing against, or when policy fills the vacuum and capture gets locked into legislation. Neither outcome is decided yet.
AI resilience is one of the largest and fastest growing coordination problem spaces we’ve seen. The demand for real solutions is outpacing the supply of credible institutions positioned to deliver them.
Why This Feels Like a Gitcoin-Shaped Problem
We built coordination infrastructure inside Ethereum for nearly a decade. Quadratic funding. Community led capital allocation. Sybil resistant identity. Programmable grant rounds. The tools and the principles were forged in one ecosystem, but the coordination failures they address are universal.
The AI transition is surfacing coordination failures at a scale and velocity we haven’t seen before. Communities need infrastructure to identify what resilience looks like for them, fund it collectively, and sustain it without depending on the institutions that created the problem. No individual retraining program, no single defensive tool, no single grant round addresses the compounding nature of the harms above.
That’s a coordination problem. And coordination infrastructure is what we’ve spent eight years learning how to build. We believe the most impactful work ahead is exporting what Ethereum public goods taught us to the world.
We’re leaning toward committing to this direction. But before we do, we want to hear from the community.
Opening the Floor
This is how we’re reading the landscape. We want to hear how the community is reading it too.
- We see a clear line of funding what matters in the past (Ethereum) and future (AI transition). Do you see this?
- What’s resonating? What’s missing?
- Which of the four harms feels most pressing from where you sit?
- Where do you think the actual leverage is?
- What does resilience mean in your work?
