Jack Dorsey, co-founder of Twitter, publicly supported venture capitalist Chamath Palihapitiya's stance advocating for open source artificial intelligence (AI), underscoring a key debate about America's strategic approach to AI development. Palihapitiya highlighted the economic peril of US policies that restrict open AI models, warning that such limitations would force American companies to pay between $26 and $56 per million tokens, while international competitors access comparable AI intelligence for as little as 50 cents to $1. This vast cost disparity threatens to cripple US firms' competitiveness on a global scale.
The crux of Palihapitiya's argument lies in the growing convergence of performance between open weight AI models and proprietary counterparts, yet with a persistent and significant price gap. Chinese AI labs have aggressively contributed to this shift, exemplified by Beijing's Moonshot AI releasing the Kimi K3 model, which recently outperformed benchmarks in coding tasks, unsettling US chip manufacturers. This development signals a narrowing capability gap between Chinese and American AI systems, intensifying the risk that US firms face structural disadvantages if open source AI is stifled.
Palihapitiya's framing links this economic challenge directly to national security. He stresses that paying exorbitant prices for AI to defend US infrastructure while adversaries operate with inexpensive, advanced models creates a strategic imbalance. Echoing this, David Sacks and researcher Sebastian Mallaby warn that the diffusion of dangerous AI capabilities is inevitable, regardless of restrictive policies. Mallaby cites models like Anthropic's Claude Mythos as examples of rapidly spreading high-level cyber capabilities, foreshadowing a future where nearly everyone might access potent AI tools.
US policymakers face a dilemma: how to control advanced AI models without ceding technological ground to competitors like China. Efforts to vet models before release aim to mitigate security risks, yet critics like Sacks argue such gatekeeping cannot contain AI development in a world where powerful systems are globally accessible. Palihapitiya and Sacks agree that restrictions may ultimately harm US economic and security interests more than openness.
Dorsey's endorsement carries additional weight given his leadership at Block, which has developed its own open source AI agent, Goose, signaling his commitment to transparency and innovation in AI. The ongoing policy decisions will shape the trajectory of US AI leadership, with implications extending to economic competitiveness and national defense.
This material is informational and does not constitute financial advice.
