Tuesday took us from Anthropic to the RAND Corporation to the Carnegie Endowment for International Peace, and it turned out to be the day that most challenged my assumptions. I walked in expecting conversations about policy. What I got instead was a series of honest reckonings with uncertainty.

The morning began at Hoover Institution's DC office, where we met with Tarun Chhabra, Head of National Security at Anthropic. Our visit came just days after Anthropic's CEO published a landmark essay calling for a global slowdown in AI development. The timing made the conversation feel urgent in a way that none of us had quite anticipated. What struck me most was how candidly the discussion addressed the question of who should be making decisions about these technologies. The answer, surprisingly, was not the companies building them. The insistence that regulation needs to be democratic, that Congress has a role to play, and that the people at the frontier of AI development should not be the ones setting the rules for everyone else was not what I expected to hear from inside the industry. It reframed the way I think about the relationship between innovation and governance.

At RAND, we sat down with Salman Ahmed, Distinguished Chair in Strategy and Diplomacy, who brought decades of experience from the UN and the White House to a conversation about the forces reshaping the international order. He described the current moment as defined by two colliding realities: the unraveling of the post-Cold War era and the rise of AI. One observation that stuck with me was the idea that the Bay Area tends to focus on change while the rest of the world focuses on continuity, and that nobody agrees on what belongs in which category. In a week full of conversations about defending democracy, RAND offered something rarer: a genuine admission that the story is still unfolding and that even the most experienced strategists are improvising.

Our final stop was Carnegie, where we met with Frances Z. Brown, Vice President for Studies. The conversation shifted from technology to the parts of the world that risk being forgotten. Africa is experiencing both the most democratic transitions and the most democratic reversals of any region right now, and yet it has been pushed off the strategic agenda by the focus on Ukraine, China, and AI. That observation was a useful corrective. The discussion also touched on the elimination of USAID and what it means for American influence abroad, reinforcing a pattern I noticed across the entire week: the institutions that support democracy are stretched thinner than ever.

For me, the thread connecting all three meetings was a question I brought with me from my own research. My thesis examines Russia's looting and destruction of Ukrainian cultural heritage, and I have been wrestling with whether cultural heritage recovery ever makes it onto the agenda when institutions plan for what comes after conflict, or whether it is always an afterthought. Tuesday confirmed my sense that it remains largely invisible in the policy conversation, which only sharpened my conviction that the work of making it visible matters.

If I could share one takeaway with an incoming honors student, it would be this: the people shaping policy in Washington are not working with settled frameworks. They are navigating enormous uncertainty in real time. And the chance to sit across from them and ask questions is the kind of experience that changes how you read everything after it.

~ Amalia Abecassis

For our first meeting of the day, we visited the Hoover Institution's outpost in DC. We met with Tarun Chhabra, a visiting Hoover Fellow and Anthropic's current head of national security policy. We learned valuable insights into how frontier model labs assess the risks of powerful AI development and how they think about aligning AI's public benefit with national security and innovation. It was amazing to hear directly about how Anthropic sees the AI landscape today and the challenges it sees going forward. We also discussed AI’s implications for democracy and how this technological wave may be an opportunity for developing countries to build infrastructure and expand social access. While I candidly felt that we as humans are only at the first step of AI safety and still haven’t identified all the potential dangers of AI development, I was comforted by the thought that frontier model labs are asking the right questions about safety and innovation. 

Next, we made our way to the Pentagon City area to speak with Salman Ahmed of RAND. He spoke broadly about his view of the state of the world and the main forces he identified as most influential today. Unsurprisingly, he spoke a bit about the risks of AI and especially about AI’s effect on the global information space. Notably, however, we heard a lot about how most people — outside of Silicon Valley, at least — are concerned about the continued transition away from the Post-Cold-War world order and how there is, internationally, little to no agreement on what comes next. It was refreshing to hear his thoughts on the other forces in the world shaping our future that are not AI. I particularly enjoyed his attention to the idea of a local political economy and how near-term effects of policy for communities can drive larger change. Salman’s thoughts prompted me to think more deeply about my own frameworks for viewing the world, especially how I identify mechanisms for change. 

In the afternoon, we had our last meeting of the day with Frances Brown and Andrew O’Donohue of the Carnegie Endowment for Peace. We received a great breakdown of current democracy policy in the US and how the meaning of democracy can shift across partisan lines depending on each’s priorities and values.

Additionally, responding to my question about the relationship between think tanks and government positions, Frances shed light on how people move back and forth and how priorities look different on each side. In particular, she talked about how work in government can raise questions for her that she then has time to return to and study in depth at a think tank. This highly permeable cross-industry relationship was not something I was previously aware of, and I really enjoyed hearing about it. After two days of meetings, I felt I was starting to pick up on aspects of the broader policy ecosystem in Washington, D.C. We heard from a variety of think tanks and government-adjacent organizations, and I could better see how these groups interacted. While each organization had its own characteristics, I was surprised by the consistent academic feel across them all. 

Overall, I was struck by how each speaker had breadth and depth of knowledge across a spectrum of topics, and by how articulate and thoughtful they were in thinking through the forces that shape our world. While I think everything is connected in some way, I felt the speakers we heard from identified the most relevant factors and defended their choices well. 

Lastly, we finished the day with a thoughtful visit to the Lincoln Memorial and reflected on the relevance of the Gettysburg Address and Second Inaugural today.

~ John Churchill