The AI Metagame: Episode 04 with Lev Milman

Episode 4 | The 600,000-Claim Brain: How Lev Milman Uses AI to Map Reality

How do you stay ahead of the curve when the "curve" is moving at the speed of light? In this episode, we sit down with Lev Milman, a former U20 US Chess Champion and macro trader, to explore his radical, AI-driven approach to information processing and belief updating.

Lev has constructed a massive digital knowledge base consisting of over 600,000 atomic claims about the world. He doesn't just "read the news"—he has built a sophisticated "information funnel" that uses Large Language Models (LLMs) to ingest, grade, and de-duplicate content from over 1,300 RSS feeds, hundreds of YouTube channels, Twitter, Discord, and academic journals.

We dive into:

• The "Greater" and the "Deduper": How Lev uses AI to filter out 80% of incoming noise and focus only on information with high novelty and belief-update potential.
• Hyper-Speed Learning: Lev’s workflow for "listening" to thousands of words of summarized research at 4.5x speed using custom text-to-speech setups, often while multitasking.
• The Chess Champion’s Edge: How a childhood spent studying chess for 10 hours a day wired his brain for the deep, contingency-based reasoning required to navigate the AI frontier.
• Actionable Intelligence: Moving from curiosity-driven learning to a practical framework where information is rated by how much it changes his decisions in fields like AI development and health.

Lev makes a compelling case that in the AI era, the ultimate competitive advantage isn't just knowing more—it's having a superior system for updating what you believe.