AI value shifts from models to distribution
Long-term financial value in the AI sector will accrue to the distribution, application, and energy/compute layers rather than the foundation Large Language Model layer.
The argument
The guests argued that LLMs face limited value capture due to intense competition, open-source alternatives, and the fact that model weights are easily replicated or licensed. Instead, companies with direct customer distribution (like Apple) or infrastructure advantages (like Microsoft, Nvidia, and energy providers) are positioned to capture the bulk of the industry's profits.
The thesis, stress-tested
✓ What validates it
- ✓Apple successfully monetizing third-party AI models through its operating system distribution
- ✓Enterprise adoption metrics showing sustained revenue growth for AI enablement services and hardware providers like Dell
▸ Risks discussed
- ▸Massive capital expenditure requirements for compute and energy could depress margins
- ▸Rapid technological shifts could render current hardware or distribution channels obsolete
Hear it yourself
"important as the system scales, as they become more institutional grade. And just similar to that trend, you know, we're seeing a lot of settlement values where it was, as we're reporting our earnings recently, we saw $21,000,000,000,000 of on chain settlement in USDC in the first quarter alone. That speaks to how the infrastructure expanding and how liquidity is improving along with, the largest platforms, whether centralized or decentralized. Alright. So pulling on that thread a little bit, if we zoom out and look at the ecosystem, a lot of USDC flows also integrate quite tightly with Coinbase."
09:45
AFFILIATE LINK · ZORTIX MAY EARN A COMMISSION · NEVER A RECOMMENDATION TO TRADE