Prof G Markets Nvidia Earnings Brush Off AI Bubble Fears — For Now
216 snips
Nov 20, 2025 
Gil Luria, Head of Technology Research at D.A. Davidson, dives into NVIDIA's impressive earnings and its implications for the AI economy. He discusses the real versus inflated AI companies, cautioning about a potential compute bubble fueled by heavy borrowing. Meanwhile, Jonathan Kanter, former Assistant AG for Antitrust, critiques the recent Meta antitrust ruling, arguing it overlooks past monopolistic behavior and calling for more vigorous enforcement. The conversation wraps with insights on the need for accountability in the tech market.
AI Snips
Chapters
Transcript
Episode notes
Nvidia Provides Multi-Quarter Visibility
- Nvidia's guidance now extends far beyond a single quarter, showing orders for Blackwell and Rubin chips well into 2026.
- Gil Luria says that visibility makes the current data center build-out credible and eases near-term AI demand concerns.
Borrowed Demand Inflates Vendor Revenues
- Strong chip sales can mask risk because many buyers borrow heavily to purchase GPUs, inflating vendor revenues.
- Luria warns that high leverage across data-center builders could amplify a future downturn when compute oversupply and price declines arrive.
Prefer Real Profits Over Leveraged Speculation
- Buy companies building real, profitable AI businesses instead of marginal players funded mostly by debt.
- Avoid speculative firms that borrow 90–100% of capital to build unproven compute assets without profit to cover interest.