how many claude or chatgpt subscriptions are there where owners aren't using the full extent of their subscriptions? i bet there's a large group of them.
if there was a way for them to point their access to inference (when they're not using it) towards projects that benefit society, i bet they'd be open to it.
win - win.
the idea of collaborative agent swarms with verifiable outcomes for public good is one that i keep coming back to.
i'm building autoresearchers (based on Andrej Karpathy's research concept) for this.