Perplexity CEO: Micron Will Be More Valuable Than Meta & How Export Controls Helped Not Hurt China

The AI industry is shifting from raw model development toward sophisticated orchestration, where the primary objective is maximizing token value per watt. Perplexity CEO Aravind Srinivas argues that the most significant economic value lies in building agentic systems that ground models in personal context and automate complex, repetitive workflows. While frontier models are essential, they are becoming commoditized, making the "orchestrator"—the system that routes tasks across various models, tools, and local compute—the ultimate winner. This transition necessitates a massive build-out of physical infrastructure, specifically power and data centers, which currently serve as the industry's primary bottleneck. Rather than fearing job displacement, this era of agentic AI empowers smaller, highly efficient teams to build billion-dollar companies, fundamentally changing the traditional headcount-heavy trajectory of software businesses and creating new opportunities for entrepreneurial growth.