Mark Zuckerberg stood at Georgetown University’s Gaston Hall in October 2019 and delivered an address that turned out to be surprisingly prescient. While his contemporaries in tech were beginning to retreat from the ideal of unfettered expression, Zuckerberg made an impassioned case for free speech as a cornerstone of technological progress. 

What seemed then like a defense of Facebook’s content policies now reads like a blueprint for Meta’s approach to artificial intelligence.

That philosophical foundation has manifested most notably in Meta’s release of Llama, an open-source AI model that represents a radical departure from the closed, proprietary systems that have dominated the field. While many of Meta’s competitors guard their models behind APIs and usage restrictions, Meta has chosen to make its technology freely available to researchers and developers worldwide.

Meta’s approach to AI isn’t mere altruism. It’s a calculated bet that an open ecosystem’s distributed innovation will ultimately outpace closed systems’ controlled development.

The numbers suggest this bet is paying off. Within months of its release, Llama 2 had been downloaded more than 30 million times, spawning countless derivatives and applications. This proliferation of innovation mirrored the early days of the internet, when open protocols and standards created the foundation for unprecedented technological growth.

The parallel isn’t accidental. Meta has invested heavily in AI infrastructure. It is building one of the world’s largest AI supercomputers and assembling a formidable research team. Unlike its competitors, who have largely treated AI development as a proprietary arms race, Meta has chosen to share its advances with the global research community. This approach harnesses what has historically been America’s most significant competitive advantage — the ability to foster and lead global networks of innovation.

Critics abound, of course. Some point to Meta’s checkered history with privacy, bias and content moderation as reasons for skepticism. Others suggest that Zuckerberg’s embrace of openness is merely a Machiavellian strategic pivot in response to recent elections.

These criticisms miss a crucial point. Meta’s past struggles with managing online discourse have given it unique and unprecedented insight into the challenges of deploying transformative digital technologies at scale.

The timing of this open-source push could not be more critical. As China accelerates its AI development, America’s traditional advantage in technological innovation faces its stiffest test yet. The U.S. government’s response has focused mainly on regulatory frameworks and potential restrictions, creating a vacuum that private sector leaders must fill.

Meta’s approach offers a compelling alternative. Making powerful AI models freely available creates a global laboratory for innovation while ensuring that development aligns with democratic values of openness and free expression. This strategy has already begun to bear fruit: Researchers have used Llama to advance everything from transforming fetal heart screening to advancing personalized medicine, often achieving results that rival or exceed those of closed systems.

The implications extend beyond technological competition. As AI becomes central to economic and military power, the question of how it develops — openly or in secret, collaboratively or competitively — takes on geopolitical significance.

Meta’s bet on openness represents not just a corporate strategy but a distinctly American vision for AI’s future: one built on the free exchange of ideas rather than centralized control.

For those uncomfortable with this vision, consider the alternative: a fragmented American AI effort hampered by excessive caution and restriction, watching as more aggressive competitors seize the initiative.

The choice isn’t whether to accelerate AI development — global competition has already made that decision for us. The choice is whether we’ll do so in a way that reinforces or undermines the values that have historically made U.S. innovation successful.

Meta’s commitment to open-source development offers a blueprint for maintaining U.S. leadership in AI while staying true to our foundational principles of openness and free expression. In the end, the future of AI may depend less on who can build the most powerful models than on who can create the most robust ecosystem for innovation.

In that race, openness is our strongest advantage. Meta’s open-source AI strategy not only echoes the much-welcomed optimism from the Trump administration but embodies the very principles of free exchange and open collaboration that have historically driven U.S. technological dominance.

It’s time to let Zuck cook.