Camp Kotok, AI, and the Questions That Matter
What a weekend at the “Shadow Fed” revealed about artificial intelligence, durable advantage, and the value of better questions.
Eric Schlesinger · August 2026
Most people have never heard of Camp Kotok, let alone have any real appreciation for how special this gathering is.
Every year, a small, by-invitation-only group of 50, gather at Leen’s Lodge in northern Maine. It is always in the same location and the same time of year. Think summer camp for an extremely unique group, including open seating dining hall, canoe fishing (two people plus a guide per boat), bunk sharing (thankfully mine didn’t snore), talent shows (panels in our case), and the mystery juice served at every meal.
One notable difference? The mystery juice. As an attendee, you are required to supply one case of fine wine or an equivalent spirit. That combination makes the conversations about markets, economics, monetary policy, geopolitics, technology, the forces that may shape what comes next, and the one that got away, that much more interesting.
This gathering is intentionally small and has historically included investors, economists, market strategists, journalists, business leaders, policymakers, and people connected to the central-banking world. The broader Global Interdependence Center community includes former Federal Reserve leaders, Federal Reserve governors, and senior officials from central banks around the world.
“The Shadow Fed.”
Meanwhile, the setting is about as far from Washington or Wall Street as you can get. Camp runs under the Chatham House Rule, which is a big part of why the conversations are so amazing. People can say what they think, challenge each other, and push on ideas without worrying that every sentence is going to show up online with their name attached to it. That rule alone changes the dynamic.
This year not only was I invited, but I also had the privilege of participating in a panel discussion on artificial intelligence with Barry Norton, moderated by Eric Hale. As expected, our discussion did exactly what we hoped for: it created more questions.
The topic of AI was ever-present, surfacing repeatedly throughout the entire weekend. After all, it intersects every topic those attending think about; capital investment, productivity, labor, national security, energy, semiconductors, markets, and economic growth. Our discussion went well beyond, “Is AI really a thing?”
Barry brought a perspective on semiconductors and the manufacturing infrastructure underlying AI that fundamentally changed how I think about the complexity of coordinating a supply chain encompassing 40 countries to produce a chip. The enormous amounts of capital being committed to AI infrastructure, whether the economics ultimately justify that investment, how quickly today’s hardware could become obsolete, what AI means for employment, and how businesses should think about adoption were all topics we discussed.
Where does durable advantage live when intelligence itself becomes increasingly commoditized?
People like Barry Ritholtz, David Kotok, Eric Hale., Barry Norton, Dave Nadig, Rick Newman, Jim Bianco and many others make Camp Kotok different because there is an expectation of different conclusions.
My biggest takeaway, and a feeling reinforced throughout the event, isn’t about AI being used to get answers. It’s that AI is helping us ask better questions. To challenge our own assumptions. To combine human judgment with machine intelligence.
As a reminder, Camp Kotok operates under the Chatham House Rule, and I won’t attribute private conversations or comments to anyone without their permission. However, the AI panel was recorded and on the record. The moderator, Eric Hale, put together an excellent followup report combining the discussion, survey, and practical examples of how people in the room are using AI.
We’re making the full report available through TaeCor for anyone who wants to go deeper.
I came back from Leens Lodge knowing three things for certain: I would very much like to be invited back, the flies in Maine are vicious, and I have an infinite number of questions and very few answers.
At this point in the development of AI, I believe the questions are far more valuable than the answers.
What are we going to do differently because intelligence is suddenly abundant to everyone?
Thank you to David Kotok, Jill Fornito and the entire Global Interdependence Center team for creating the space where these messy, critical, and necessary conversations can happen.
The data from the weekend is still being synthesized, but the human insights remain the most valuable takeaway from the Shadow Fed.
At this point in the development of AI, I believe the questions are far more valuable than the answers.