Draft Moskovitz: The Best Last Hope for Constructive AI Governance

A viral essay on LessWrong by Oliver Kuperman makes a urgent case for Dustin Moskovitz, the Asana co-founder and prominent AI safety philanthropist, to run for U.S. President in the 2026 election. The core argument is that this may be one of the last elections capable of meaningfully shaping AI governance before the potential development of Artificial General Intelligence (AGI). Kuperman contends that the current "race to the bottom" dynamics among AI labs like OpenAI and Anthropic—where competitive pressure discourages safety precautions—make robust government regulation essential, as corporate self-regulation is insufficient.

The essay builds its case by highlighting the distorted incentives in AI development, where being a first-mover could confer overwhelming economic and strategic power, pushing companies to "move fast and cut corners." It extensively quotes leading AI scientists like Geoffrey Hinton and Yoshua Bengio, who have publicly stated that market forces alone cannot mitigate catastrophic risks and that government action is required. Kuperman argues that a Moskovitz presidency would prioritize safety, but even a failed campaign would succeed by thrusting AI safety into the national spotlight, influencing policy debates, and catalyzing a dedicated political movement around the issue.