Plain English with Derek Thompson Everybody Thinks AI Is a Bubble. What If They’re Wrong?
547 snips
Oct 17, 2025 
Azeem Azhar, an investor and author of the Exponential View newsletter, dives deep into the debate surrounding AI as a financial bubble. He argues against conventional wisdom, suggesting that today’s AI differs fundamentally from past bubbles due to real user demand. Azeem discusses the implications of rapid GPU depreciation, the strain of data center spending, and the potential for AGI to reshape the economy. He also outlines critical revenue growth needed for sustainability, questioning whether AI is truly a bubble or simply misunderstood.
AI Snips
Chapters
Transcript
Episode notes
Strict Two-Part Definition Of A Bubble
- A bubble requires both a deep market correction and a sustained drop in productive capital investment.
- Azeem defines thresholds: >50% market decline and ~50% fall in productive capital for multiple years.
Dot‑Com Memories Highlight Key Difference
- Azeem recalls the dot-com era when founders literally climbed fire escapes to pitch investments.
- He contrasts empty 2000s sites with today's widely used generative AI tools.
Real Usage Separates AI From 2000 Bubble
- Generative AI adoption is real and fast, with billions in spending and consumers using tools daily.
- That real usage differentiates today's cycle from 2000 where many products had no users.