In a strategic pivot for the AI industry, Baidu CEO Robin Li has challenged the prevailing hardware-centric narrative by introducing Daily Active Agents (DAA) as the ultimate metric for AI value. While hardware giants like Nvidia emphasize Token generation and throughput, Li argues that the true measure of success lies in the practical application and business value delivered by active AI entities.
This shift represents Baidu’s strategic evolution from training foundational models to deploying enterprise-grade applications. By prioritizing DAA over Tokens, Li highlights a fundamental transition: moving from measuring how much energy or compute is consumed to measuring how much productivity is generated. This metric focuses on the tangible execution of tasks and the resolution of complex business problems by AI Agents.
The contrast between Token-centric hardware views and DAA-centric application views marks a new era in AI competition. As the ecosystem matures, the emphasis is no longer on raw technical power but on the deployment of intelligent agents that can drive real-world outcomes. This philosophy advocates for a sustainable AI economy where utility and user engagement become the primary drivers of technological progress.