AI changes the stakes.
AI companies can scale influence and wealth faster than public institutions can adapt: compute leverage, deployment speed, labor-market effects, safety externalities, and governance gaps can all compound quickly. If builders capture extraordinary upside from that shift, it is worth deciding early that some of it will flow back to public-good work.
decide while incentives are still clean.
This is not asking founders to give before they have liquidity. It is asking them to set the rule before the outcome is known, while the decision is still principled rather than retrospective.
the destination stays founder-directed.
The commitment is not a funnel to one organization. Founders can later route proceeds to eligible work that fits the actual risks and opportunities of the AI era.
safety and evals
Technical safety, evaluations, red-teaming, interpretability, and safer deployment infrastructure.
governance capacity
Policy talent, public institutions, accountability work, and infrastructure for better decisions.
education and transition
Learning access, child development, worker transition, and equitable adaptation to AI.
public-interest tech
Open tools, civic resilience, local infrastructure, and technology serving communities.