We won't be able to deploy a torrent of funding even if we multiply grantors a thousand-fold. The reason: we don't have the talent or ideas to leverage it. Maybe automated research will one fill that gap, then request all our funding to run its experiments. But until then, the best way around the grantor bottleneck is to start gifting. 

Why grants are the wrong tool for this moment 

Manifund has diagnosed the problem clearly: grants are budgeted to cost rather than value, they're restrictive, they're slow, and they create funder chicken. Grant funding today is a command economy, and command economies are bad at exactly the thing we need right now — surfacing the people and ideas we don't yet know exist.

The deeper issue is that grants assume the talent pipeline already exists. They route resources to researchers who've already identified themselves as researchers, written proposals, and cleared the signaling cost of looking fundable. That's fine when the bottleneck is money. It's the wrong model when the bottleneck is the pipeline itself — when we're trying to find out who could do the work if they had access to the tools.

The proposal: buy the maker space

Instead of finding more grantors to funnel the incoming funding, we should convert a meaningful fraction of funding into gifts — durable, specialized infrastructure that anyone can use. Think of it as a maker space for AI-era research: a public stock of tools, too expensive or specialized for individuals to own, that people can check out on demand.

Concretely, torrents-of-funding recipients might purchase or commission public goods that currently sit behind institutional walls. Some examples:

• Specialized language models fine-tuned on licensed proprietary corpora — materials science, legal archives, clinical data, engineering documentation
• Compute pools with the kind of hardware that sits idle most of the time at individual labs - up to and including quantum computers
• Curated datasets released under public-access terms, bought outright from the rights holders.

Access would be distributed like a library — or more precisely, like a tool library or maker space, since the tools can be rival and usage is finite. Members would check out time on a specialized model, return it, renew it, or queue. Some instruments would be general-access; some would require demonstrated proficiency, the same way a woodshop gates its table saw. The takeaway is that in this model, committees don't pick winning projects; instead, members choose what to pull off the shelf. 

What this moves, and what it doesn't solve

This shifts the allocation problem rather than eliminating it. Someone still decides which assets to purchase and stock. But choosing infrastructure is a wider-bandwidth decision than choosing projects: a single well-chosen asset serves thousands of users across projects the funder never imagined. 

To be very clear, this proposal doesn't replace grants; it complements them. Grants remain the right instrument for well-scoped projects with identified teams. The gifts economy sits underneath, decreasing the operational cost of those who are grant-ready, and widening the base of people who can become grant-ready in the first place. 

Why now

The scariest version of the funding torrent isn't waste. It's that funders, under pressure to deploy, route more money through the same narrow committees to the same recognizable recipients — and we end up with a concentration of capability around whoever was already legible to grant-makers in 2025. A gifts economy pushes in the other direction. It converts a moment of financial abundance into wide public capacity that outlives any particular funder's priorities.

A brief epilogue

Global warming has introduced torrents of rain into my neighborhood. My home, and my family members' homes, have been flooded more in the last five years than in the last twenty-five years. The problem is well known: our sewage pipes are too narrow.

Which made me think of an initiative in Denmark. Citizens started removing squares of concrete from their sidewalks and replacing them with plants. This had a great aesthetic effect on the city, but it also played a functional role: floods and sewer stress decreased. By adding soil area, the torrents could be absorbed and used to grow flowers.

That's how I see this proposal: it increases soil area. Public goods absorb rather than funnel torrents of funding. Less goes to waste, and sometimes, beautiful things grow — for the benefit and safety of all.