Coordinating Collective Action for AI Safety
"Show me the incentives, I'll show you the outcome."
Why do people notice the same problems independently but do nothing? Why do companies cut corners they'd rather not cut? Why do regulators keep getting outmaneuvered?
What can seem like failures of ethics or goodwill are, in fact, collective action and coordination problems—well-studied failure modes that show up everywhere, from climate policy to whistleblowing to AI governance. Luckily, they have solutions.
Join us at Mox for a 90-minute working session on the coordination problems shaping the most important technology of our lifetime, and what it actually takes to solve them.
Agenda
The mechanics of collective action and coordination problems: what they are, how they differ, why smart and well-intentioned people get trapped by them, and the recurring patterns from labor organizing to pandemic response.
The AI safety angle: capability arms races, safety research free-riding, regulatory arbitrage, critical information hoarding, and the systematic neutralization of would-be whistleblowers.
An interactive coordination game: experience an incentive trap firsthand and watch how small changes to the rules change the equilibrium.
Group brainstorm: surface the coordination problems attendees find most pressing and workshop on what it would take to move them.
Bring a problem you're stuck on. Leave with a clearer map of why it's stuck and the leverage points that could shift it.
To enter Mox, use the URL https://moxsf.com/door?evt=18PXYXb7d4CE. This code will work from 5 PM – 6:30 PM.
Facilitator
Jordan Braunstein is the founder of Spartacus.app, a coordination platform for kickstarting collective action and an ACX grant recipient. He holds a B.S. in behavioral economics and brings 15 years of experience in growth and product management across early-stage startups.