Deep Dive: How Machines Are Becoming Better Investors Than Humans
Chamath Palihapitiya
· Chamath Palihapitiya
· February 19, 2026 at 15:00
· ⏱ 6 min read
| Read on Substack ↗
Summary
=== SUMMARY ===
•The primary source of competitive advantage in financial markets has irrevocably shifted from human-centric strategies (judgment, relationships, conviction) to autonomous, code-driven systems that can process vast datasets and execute with speed and discipline.
•The evolution of investing is progressing through five levels, culminating in "Agentic AI Investing," where fully autonomous systems can plan, act, and self-correct. This represents a long-term, structural shift in the asset management industry.
Summary
Humans cannot compete with machines in modern investing because information is instantly priced and markets move in milliseconds. The article argues that the best long-term returns come from code-driven funds like Renaissance's Medallion (39% annualized), and that autonomous investing — from algorithmic to agentic AI — is the only sustainable competitive advantage, accelerating the decline of traditional active management.
•Renaissance Medallion Fund earned ~39% net annual returns since the late 1980s, compared to S&P 500's ~10% and Berkshire Hathaway's ~19.9%.
•Three key breakthroughs enabled automated investing: Markowitz's portfolio theory (1952), Sharpe's CAPM (1964), and Black-Scholes options pricing (1973).
•Medallion's edge comes from executing tiny statistical advantages across thousands of instruments simultaneously, a scale unachievable by humans.
•The article defines five levels of investing automation: Manual, Algorithmic, Automated, Autonomous, and Agentic AI, with Medallion operating at Levels 4-5.
•In October 2025, seven frontier AI models each managed $10,000 in real capital with zero human intervention at Alpha Arena, demonstrating the current state of Agentic AI investing.
•The article is an educational deep dive, not an investment recommendation or personal trade disclosure.