Perplexity – Frontier AI systems are closing the gap between model intelligence and real-world utility. New models, compute architectures, and orchestration patterns are enabling these systems to accomplish tasks deemed impossible just a few months ago. This rapid innovation has proved a boon to AI users by magnifying their leverage and agency. Yet it has also created a lag between the technological frontier and our understanding of precisely how knowledge work is evolving in response. How does frontier AI change the nature of knowledge work across professions? Which structural and economic transformations in this work might we expect? We seek to close this gap through careful empirical analysis of Perplexity usage. Today, in collaboration with Harvard Business School researchers, we’re sharing our first comprehensive study of Perplexity Computer in real-world deployment. Our findings suggest that Computer expands both the breadth and depth of what users can accomplish, at lower cost. Computer users are getting more done, working at higher levels of abstraction, and crossing disciplinary boundaries to unlock value outside their own professions. This article presents the highlights from our study. Detailed methodology and findings are available in our technical report.”