The paradox of AI accelerationism and the promise of public interest AI

Abstract

Over the past 3 years, the idea that artificial intelligence (AI) development should be accelerated without restraint to drive radical societal change-a phenomenon known as AI accelerationism-has gained traction in Silicon Valley. The most prominent form of this belief is known as "effective accelerationism," and adherents view unrestricted progress in AI as a solution to humanity's problems. Rooted in the broader political ideology of accelerationism, this movement positions itself directly in opposition to "AI doomers" or "decelerationists," who advocate for AI safety and slow, cautious development. Effective accelerationists often dismiss concerns about existential risk, arguing that technological stagnation is a greater danger than uncontrolled advancement. Public interest AI inverts AI accelerationism, promoting more equitable, sustainable, and democratically accountable technological progress. This approach directly addresses the risks and potential harms of unregulated acceleration by aligning AI with public values rather than solely with private interests.