Lenovo CEO Proposes 'Private Data Flywheel' for Hybrid AI Shift
At the 2026 Lenovo Venture CEO Annual Conference, Lenovo CEO Yang Yuanqing introduced the concept of the 'private data flywheel,' arguing that AI industry competition is shifting from model parameters to hybrid AI deployment capabilities. This perspective highlights a broader industry trend where public data commoditizes base models, making proprietary private data and edge-to-cloud integration the critical differentiators for enterprise and personal AI utility. Yang noted that Lenovo's AI PC holds a 25% global market share, with personal smart agents upgraded to version 4.5, and predicted that 70% to 80% of future AI tokens will be generated at the edge or on device.
## BACKGROUND
Hybrid AI is an architecture that combines edge computing on local devices with cloud-based processing to balance performance, privacy, and responsiveness. A data flywheel refers to a self-reinforcing loop where increased product usage generates more data, which in turn improves the AI model to attract even more users.