← Home
Invest Like the Best · April 23, 2026 · 45m

Dylan Patel - The Infinite Demand for Tokens, Claude Mythos, and Supply Constraints

Dylan Patel
Show Notes

Dylan Patel, founder of SemiAnalysis, discusses the explosive demand for frontier AI models and the supply constraints limiting their scaling. He explains why frontier models command nearly unbounded willingness to pay, how his firm's AI spending has grown from tens of thousands to $7M in a year, and the critical bottlenecks in memory, logic, and fab equipment that will determine how fast AI infrastructure can scale. The conversation also covers perception challenges facing leading AI labs.

This summary was generated from show notes and public descriptions, not from a full transcript review. Details may contain inaccuracies.

Highlights

•
Frontier Model Dominance and Unbounded Willingness to Pay
Only the frontier model—the leading cutting-edge AI system—commands meaningful enterprise demand and willingness to pay, regardless of price.
ClaudeChatGPTGemini
•
AI Infrastructure Spending Acceleration as Demand Signal
SemiAnalysis' own AI spending grew from tens of thousands of dollars to $7 million in one year, demonstrating explosive enterprise demand for frontier models.
ClaudeChatGPTGemini
•
Supply Constraints Across Memory, Logic, and Fab Equipment
Three critical bottlenecks will constrain AI scaling: high-bandwidth memory (HBM), logic chips, and the manufacturing equipment needed to produce them.
ClaudeChatGPTGemini

Editorial

•
Public Perception and Credibility Crisis at Leading AI Labs
Leading AI labs face growing public perception problems that require strategic communication shifts.
ClaudeChatGPTGemini

Misc

✧SemiAnalysis' AI spending jumped from tens of thousands to $7M in one year—a massive signal of enterprise demand acceleration
✧Frontier model dominance: companies are willing to pay nearly unbounded prices for access to the leading model
Was this useful?