The Dynamist

During the pandemic from 2020 to 2021, Congress dropped $190 billion to help reopen schools, provide tutoring, and assist with remote learning. The results? Fourth graders' math scores have plummeted 18 points from 2019-2023, eighth graders’ have dropped 27 points - the worst decline since testing began in 1995. Adult literacy is deteriorating too, with Americans in the lowest literacy tier jumping from 19% to 28% in just six years. Are we watching the collapse of academic achievement in real time?In this episode, education policy veteran Chester Finn joins us to examine this crisis and potential solutions. Drawing on his experience as a Reagan administration official and decades of education reform work, Finn discusses why accountability measures have broken down, whether school choice can right the ship, and if the federal government's education R&D enterprise is fixable. Joining the conversation are FAI's Dan Lips and Robert Bellafiore, who recently authored new work on leveraging education R&D to help address America's learning challenges.This is part one of a two-part series examining the state of American education and exploring paths forward as a new administration takes office with ambitious - and controversial - plans for reform. 

The newly established Department of Government Efficiency (DOGE) has put state capacity back in the spotlight, reigniting debates over whether the federal government is fundamentally broken or just mismanaged. With Elon Musk at the helm, DOGE has already taken drastic actions, from shutting down USAID to slashing bureaucratic redundancies. Supporters argue this is the disruption Washington needs; critics warn it’s a reckless power grab that could erode public accountability. But regardless of where you stand, one thing is clear: the ability of the U.S. government to execute policy is now under scrutiny like never before.That’s exactly the question at the heart of this week’s episode. From the Navy’s struggles to build ships to the Department of Education’s FAFSA disaster, our conversation lays out why the government seems incapable of delivering even on its own priorities. It’s not just about money or political will—it’s about outdated hiring rules, a culture of proceduralism over action, and a bureaucracy designed to say "no" instead of "go." These failures aren’t accidental; they’re baked into how the system currently operates. Jennifer Pahlka, former U.S. Deputy Chief Technology Officer under President Obama and Senior Fellow at Niskanen Center and Andrew Greenway, co-founder of Public Digital, join.The solution? A fundamental shift in how government works—not just at the leadership level, but deep within agencies themselves. She advocates for cutting procedural bloat, giving civil servants the authority to make real decisions, and modernizing digital infrastructure to allow for rapid adaptation. Reform, she argues, isn’t about breaking government down; it’s about making it function like a system designed for the 21st century. Whether DOGE is a step in that direction or a warning sign of what happens when frustration meets executive power remains to be seen.

At Trump's second inauguration, one of the biggest stories, if not the biggest, was the front-row presence of Big Tech CEOs like Google’s Sundar Pichai and Meta’s Mark Zuckerberg—placed even ahead of Cabinet members. As the plum seating signaled a striking shift in Silicon Valley's relationship with Washington, just 24 hours later, the administration announced Stargate, a $500 billion partnership with OpenAI, Oracle, and other tech giants to build AI infrastructure across America.But beneath the spectacle of billionaire CEOs at state functions lies a deeper question about the "Little Tech" movement—startups and smaller companies pushing for open standards, fair competition rules, and the right to innovate without being crushed by either regulatory costs or Big Tech copycats. As China pours resources into AI and semiconductors, American tech policy faces competing pressures: Trump promises business-friendly deregulation while potentially expanding export controls and antitrust enforcement against the very tech giants courting his favor.To explore this complex new paradigm, Evan and FAI Senior Fellow Jon Askonas are joined by Garry Tan, CEO of Y Combinator, the startup accelerator behind Airbnb, DoorDash, and other alumni. As both a successful founder and venture capitalist, Tan discusses what policies could help startups thrive without dipping into overregulation, and whether Silicon Valley's traditionally progressive culture can adapt to Trump's tech alliances. You can read more about YC’s engagement with Washington, DC here.

Chinese AI startup DeepSeek’s release of AI reasoning model R1 sent NVIDIA and other tech  stocks tumbling yesterday as investors questioned whether U.S. companies were spending too much on AI development. That’s because DeepSeek claims it made this model for only $6 million, a fraction of the hundreds of millions that OpenAI spent making o1, its nearest competitor. Any news coming out of China should be viewed with appropriate skepticism, but R1 nonetheless challenges the conventional American wisdom about AI development—massive computing power and unprecedented investment will maintain U.S. AI supremacy.The timing couldn't be more relevant. Just last week, President Trump unveiled Stargate, a $500 billion public-private partnership with OpenAI, Oracle, SoftBank, and Emirati investment firm MGX aimed at building AI infrastructure across America. Meanwhile, U.S. efforts to preserve its technological advantage through export controls face mounting challenges and skepticism. If Chinese companies can innovate despite restrictions on advanced AI chips, should the U.S. rethink its approach?To make sense of these developments and their implications for U.S. technological leadership, Evan is joined by Tim Fist, Senior Technology Fellow at the Institute for Progress, a think tank focused on accelerating scientific, technological, and industrial progress, and FAI Senior Economist Sam Hammond. 

As revelations about Meta's use of pirated books for AI training send shockwaves through the tech industry, the battle over copyright and AI reaches a critical juncture. In this final episode of The Dynamist's series on AI and copyright, Evan is joined by FAI's Senior Fellow Tim Hwang and  Tech Policy Manager Joshua Levine to discuss how these legal battles could determine whether world-leading AI development happens in Silicon Valley or Shenzhen.The conversation examines the implications of Meta's recently exposed use of Library Genesis - a shadow library of pirated books - to train its LLaMA models, highlighting the desperate measures even tech giants will take to source training data. This scandal crystallizes a core tension: U.S. companies face mounting copyright challenges while Chinese competitors can freely use these same materials without fear of legal repercussions. The discussion delves into potential policy solutions, from expanding fair use doctrine to creating new statutory licensing frameworks, that could help American AI development remain competitive while respecting creator rights.Drawing on historical parallels from past technological disruptions like Napster and Google Books, the guests explore how market-based solutions and policy innovation could help resolve these conflicts. As courts weigh major decisions in cases involving OpenAI, Anthropic, and others in 2024, the episode frames copyright not just as a domestic policy issue, but as a key factor in national technological competitiveness. What's at stake isn't just compensation for creators, but whether IP disputes could cede AI leadership to nations with fewer or no constraints on training data.

In the third installment of The Dynamist's series exploring AI and copyright, FAI Senior Fellow Tim Hwang leads a forward-looking discussion about how market dynamics, technological solutions, and geopolitics could reshape today's contentious battles over AI training data. Joined by Jason Zhao, co-founder of Story AI, and Jamil Jaffer, Executive Director of the National Security Institute at George Mason University, the conversation moves beyond current lawsuits to examine practical paths forward.The discussion challenges assumptions about who really stands to gain or lose in the AI copyright debate. Rather than a simple creator-versus-tech narrative, Zhao highlights how some creators and talents have embraced AI while others have shown resistance and skepticism. Through Story's blockchain-based marketplace, he envisions a world where creators can directly monetize their IP for AI training without going through traditional gatekeepers. Jaffer brings a crucial national security perspective, emphasizing how over-regulation of AI training data could threaten American technological leadership - particularly as the EU prepares to implement strict new AI rules that could effectively set global standards.Looking ahead to 2025, both guests express optimism about market-based and technological solutions winning out over heavy-handed regulation. They draw parallels to how innovations like Spotify and YouTube's Content ID ultimately resolved earlier digital disruptions. However, they warn that the US must carefully balance innovation and IP protection to maintain its AI edge, especially as competitors like China take a more permissive approach to training data. The episode frames copyright not just as a domestic policy issue, but as a key factor in national competitiveness and security in the AI era.

From the SAG-AFTRA picket lines to the New York Times lawsuit against OpenAI, the battle over AI's role in creative industries is heating up. In this second episode of The Dynamist's series on AI and copyright, we dive into the messy reality of how artists, creators, and tech companies are navigating this rapidly evolving landscape.Our guests bring unique perspectives to this complex debate: Mike Masnick, CEO of Techdirt, who's been chronicling the intersection of tech and copyright for decades; Alex Winter, the filmmaker and actor known for Bill & Ted's Excellent Adventure, who offers boots-on-the-ground insight from his involvement in recent Hollywood labor negotiations; and Tim Hwang, Senior Fellow at FAI, who explores how current legal battles could shape AI's future.The conversation covers everything from "shakedown" licensing deals between AI companies and media outlets to existential questions about artistic value in an AI age. While the guests acknowledge valid concerns about worker protection and fair compensation, they challenge the notion that restricting AI development through copyright law is either practical or beneficial. Drawing parallels to past technological disruptions like Napster, they explore how industries might adapt rather than resist change while still protecting creators' interests.

Copyright law and artificial intelligence are on a collision course, with major implications for the future of AI development, research, and innovation. In this first episode of The Dynamist's four-part series exploring AI and copyright, we're joined by Professor Pamela Samuelson of Berkeley Law, a pioneering scholar in intellectual property law and a leading voice on copyright in the digital age. FAI Senior Fellow Tim Hwang guest hosts. The conversation covers the wave of recent lawsuits against AI companies, including The New York Times suit against OpenAI and litigation facing Anthropic, NVIDIA, Microsoft, and others. These cases center on two key issues: the legality of using copyrighted materials as training data and the potential for AI models to reproduce copyrighted content. Professor Samuelson breaks down the complex legal landscape, explaining how different types of media (books, music, software) might fare differently under copyright law due to industry structure and existing precedent.Drawing on historical parallels from photocopying to the Betamax case, Professor Samuelson provides crucial context for understanding today's AI copyright battles. She discusses how courts have historically balanced innovation with copyright protection, and what that might mean for AI's future. With several major decisions expected in the coming months, including potential summary judgments, these cases could reshape the AI landscape - particularly for startups and research institutions that lack the resources of major tech companies.

As we approach the three-year mark of the war in Ukraine, and conflict continues to rage in the Middle East, technology has played a key role in these arenas—from cyber attacks and drones to propaganda efforts over social media. In Ukraine, SpaceX’s Starlink has blurred the lines between commercial and military communications, with the satellite broadband service supporting the Ukrainian army while becoming a target for signal jamming by Russia. What can we learn from these conflicts in Europe and the Middle East? What role will cyber and disinformation operations play in future wars? What has Ukraine taught us about the U.S. defense industrial base and defense technology? As China increases its aggression toward Taiwan and elsewhere in the Indo-Pacific, how will technology play a role in either deterring a conflict or deciding its outcome? Evan is joined by Kevin B. Kennedy,  a recently retired United States Air Force lieutenant general who last served as commander of the Sixteenth Air Force. He previously served as Director for Operations at U.S. Cyber Command.

Luke Hogg, Director of Policy and Outreach at the Foundation for American Innovation, and Josh Levine, Tech Policy Manager at the same organization, dive into the significant twists in tech policy throughout 2024. They discuss the ramifications of antitrust lawsuits against Big Tech and emerging data privacy laws aimed at protecting kids online. The duo contrasts U.S. regulations with Europe’s aggressive tech frameworks, exploring the future of AI regulation and potential shifts post-election. Tune in for their insights on what lies ahead in 2025!