Aravind Srinivas argues the future of AI isn't about building bigger data centers, it's about squeezing more value from every watt
Perplexity CEO Aravind Srinivas has a new favorite metric for the AI industry: “token value per watt per user.”
In a CNBC interview on June 3, Srinivas laid out a thesis that the winners in AI won’t be the ones who spend the most on infrastructure. They’ll be the ones who extract the most economic value from the energy they consume.
The power user problem #
For Perplexity, this dynamic is playing out at meaningful scale. The company reported approximately 45 million monthly active users as of early 2026, processing around 780 million queries each month. That’s roughly 17 queries per user per month on average, but averages are deceiving. The real compute load is concentrated among the heaviest users. Srinivas framed the competitive landscape around five variables that need to be balanced simultaneously: accuracy, latency, cost, privacy, and intelligence. Companies that can optimize across all five dimensions per unit of energy consumed will, in his view, dominate the next phase of AI competition.
A $21 billion bet on efficiency over scale #
Perplexity AI, founded in August 2022, has reached a $21.21 billion valuation after its Series E-6 funding round.
The company’s infrastructure strategy reflects the efficiency-first philosophy Srinivas is preaching. Rather than building its own data centers, Perplexity locked in a three-year, $750 million partnership with Microsoft Azure for GPU resources. That’s a deliberate choice to rent rather than own, keeping the company asset-light while still securing access to serious compute power.
Srinivas is essentially arguing that the market will eventually reward efficiency over raw capacity. If he’s right, companies whose market narratives depend on infrastructure spending could face uncomfortable recalibrations, as spending isn’t a moat if someone else can deliver the same output with fewer resources.
The $10 trillion question nobody wants to answer #
Srinivas described the potential shift toward running AI models locally on consumer hardware as a “$10 trillion question,” framing it as a potentially disruptive force against the entire centralized data center model.
For Perplexity specifically, this creates a strategic tension. The company’s core product is an AI-powered search and answer engine that currently relies on cloud infrastructure. The company has already started moving toward more complex, agentic workflows with products like “Computer,” which is designed to automate multi-step tasks rather than just answer questions. Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our