An open catalog of AI tools, networks, and capital — for the work the world actually needs.
A small group has decided AI's purpose is to consolidate power, automate labor, and target adversaries. That is one future. It is not the only one. The same models that draft surveillance dossiers can draft restoration plans. The same compute that trains autonomous weapons can train diagnostic tools for clinics that have never seen one.
"AI must be used this way" is a procurement argument wearing a philosopher's robe. There is no law of physics requiring Palantir's roadmap. There is only capital, and capital can be redirected.
Most people with world-improving ideas never find the model, the dataset, the mentor, or the $50K that would let them ship. The defense-tech pipeline is overbuilt; this one barely exists.
If you can find the right tool, the right network, and the right capital in one place — and if you can see who else is already working on it — you can start tomorrow. That is what we are building.
The dominant narrative is grey: extraction, defense, optimization. We refuse the grey. The catalog is biased toward beauty, repair, and the things humans actually need.
Seven domains, three tiers in each: ideas, projects in development, and the polished work we showcase. The AI tools, communities, datasets, and funding sources we vet sit alongside the raw ideas being shared — so builders find each other earlier.
Models, datasets, and capital for restoration, biodiversity, and decarbonization.
Explore the domain →Diagnostics for underserved clinics, mental-health access, eldercare without surveillance.
Explore the domain →Open-access learning, multilingual tutors, tools for neurodivergent learners.
Explore the domain →Legal aid, accountability journalism, voting access, public-records tooling.
Explore the domain →Smallholder agriculture, food sovereignty, water systems, land stewardship.
Explore the domain →Community land trusts, energy retrofits, public space, transit equity.
Explore the domain →Tools, standards, and infrastructure for protecting human-made music, art, language, and craft — from training-data extraction, deceptive substitution, and the quiet erosion of attribution. The manifesto's anti-extraction thesis rendered into operational practice.
Explore the Preservation domain →Different audiences, different curation. Both feed the same directory — pick the door that fits what you've got.
Mid-build or already shipped. Tell us what it does, who it serves, and where it fits.
We vet for fit, evidence, and commons-orientation. We may reach out for clarifying details before listing.
Featured in the relevant domain catalog. Standout work earns investor introductions, capital matching, and editorial features through AIPIA.
Anything you built with AI — for your community, your kids' school, your block, your weird passion project. No minimum bar of seriousness.
No editorial curation — just a fast spam check. Approved by email within a few days.
Listed in the makers gallery for others to find. Standout makes can surface in the domain catalog too.
A directory's values are defined as much by what it excludes as what it lists. The list below is not a soft preference. It is the line.
If your work is in any of these categories, this is not the directory for you. There are well-funded ones elsewhere.
The catalog opens for submissions in Q3 2026. Join the waitlist for early access, or submit a project now if you're already building one of the seven domains. If you made something small or personal, share it in the makers gallery →
// Waitlist
First edition of the directory, founding cohort invitation, and a heads-up when the catalog opens for submissions.
// Project submission
A project mid-build, or something already shipped, that fits one of the seven domains. Editorial review before inclusion. — If you made something small or personal, share it in the makers gallery →