
Merryn Talks Money Gold, Debt and the AI Boom: A Financial Historian’s Warning
29 snips
Mar 9, 2026 Edward Chancellor, investment strategist and financial historian, offers a historical take on markets, debt and gold. He discusses the AI boom and whether current enthusiasm is a bubble. He explores energy constraints, rising long-term rates and sovereign debt risks. He makes the case for gold, commodities, Japan and emerging markets over bonds and expensive growth stocks.
AI Snips
Chapters
Books
Transcript
Episode notes
AI Hype Overlooks Limits Of LLM Architecture
- AI hype contains a core truth but overlooks structural limits in model architecture that cause hallucinations.
- Edward Chancellor cites Turing Award winners and Michael Wooldridge arguing LLMs guess probabilistically and inherently misreason, limiting transformative potential.
Bubbles Build Real Infrastructure But Wipe Out Early Investors
- Technology booms have a justified societal impact but often destroy investor capital in the build phase due to overinvestment in infrastructure.
- Chancellor compares 1990s internet overcapacity (Global Crossing) to current AI infrastructure spending that may earn marginal returns.
Avoid Chasing Private AI Unicorns As You Already Own Exposure
- Individual investors already have indirect AI exposure via broad market caps and need not chase private AI winners.
- Chancellor notes JPMorgan found ~45–47% of US market had AI exposure and IPOs could push that to 55–60%, absorbed by passive funds.






