Is AI Really in a Bubble? Nvidia's Jensen Huang Says Not So Fast

While the internet is buzzing with bubble warnings—thanks partly to Michael Burry’s doomsaying—Nvidia’s CEO Jensen Huang just dropped some facts during earnings that suggest the AI boom is just getting started.

Three Platform Shifts Happening Right Now

Huang broke down what he sees as the real story:

1. CPU → GPU transition: The entire cloud computing world (hundreds of billions in spend) is shifting from traditional processors to GPUs that can handle parallel processing. This is just the beginning.

2. Classical ML → Generative AI: ChatGPT-style models are already reshaping search, recommendations, ads, and content moderation. Meta’s a perfect example—they saw a 5% conversion lift on Instagram and 3% on Facebook just by integrating gen AI into their ad targeting.

3. Agentic AI: The next frontier. Think Tesla’s Full Self-Driving or AI legal assistants like Harvey. These systems make decisions independently at scale.

Huang’s take: accelerated computing is foundational, generative AI is transformational, and agentic AI will be revolutionary.

The Catch? Nobody Actually Knows

Here’s the honest part—some data screams bubble (Tesla and Palantir’s insane valuations), while other data shows legitimate AI infrastructure buildout. The future’s unpredictable, and trying to time it usually backfires.

The smart move? Dollar-cost average into quality names rather than panic-buying or panic-selling. History rhymes but doesn’t repeat.