Orchestrators capture more value than model builders
The guest argued that AI orchestrators—which route tasks across local and server-side models—will capture the most long-term economic value in AI by optimizing for cost, privacy, and intelligence.
The argument
Pure server-side frontier models are too expensive to run continuously for 24/7 agents. By utilizing local device compute and only calling server-side models when necessary, orchestrators insulate themselves from model depreciation and benefit from improvements across the entire stack.
The thesis, stress-tested
✓ What validates it
- ✓Orchestration platforms showing accelerated revenue growth relative to underlying API spend
- ✓Widespread adoption of local, continuously learning models on consumer devices
▸ Risks discussed
- ▸Users may prefer vertically integrated first-party operating system solutions
- ▸Local hardware limitations could restrict agent capabilities
Hear it yourself
"The most exciting part for me, they're heavily investing in agentic finance. If you're scaling globally, you need a banking and finance platform that's borderless, real time, and intelligent. Check out airwallets today and see how they're helping thousands of businesses like Canva, McLaren, and Deal. Scale at airwallex.com/20bc. Terms and conditions apply. Your money is a safeguarded, not FSCS protected. See airwallex.com for more details. While airwallets helps your money move globally, Vanta helps your security keep up. What's one thing in business that's spreading as fast as AI? AI risk."
AFFILIATE LINK · ZORTIX MAY EARN A COMMISSION · NEVER A RECOMMENDATION TO TRADE