Google CEO Sundar Pichai, Google DeepMind CEO Demis Hassabis, and Google senior vice president James Manyika jointly outlined a shift in the competitive landscape for artificial intelligence, arguing that deployment speed, task reorganization, capability gaps, and physical integration will determine outcomes over the next five years.
The executives said the era of comparing model strength is ending, and the critical challenge is integrating AI into everyday workflows. Pichai said AI has moved beyond experimental tools used by a few large firms to technologies entering routine work across sectors. He cited examples, including doctors using AI to organize medical records, farmers applying AI to inform agricultural decisions, and students using AI to address learning gaps, framing value as increasingly tied to role-specific applications rather than to cloud potential alone.
Pichai disclosed that Google Cloud's backlog of orders has doubled in a year to US$240 billion, presenting this as market validation of enterprise commitments to AI services and long-term contracts. He characterized such large upfront investments as a signal that companies are shifting from asking whether to use AI to treating adoption as a necessity. According to Pichai, this purchase behavior reflects AI's transition from potential to operational transformation.
Manyika focused on the workforce implications, describing the principal uncertainty as the pace of task reorganization rather than large-scale job elimination. He argued that many job titles will remain, but that task distribution is changing: routine portions of roles, such as repetitive customer service inquiries, are increasingly automated, while remaining human work centers on complaints, escalations, and retention efforts requiring judgment. Manyika emphasized continuous learning and rapid upskilling as baseline requirements for workers and a determinant of organizational competitiveness.
Hassabis framed the near future as an era of AI agents capable of proactive task execution rather than reactive responses to prompts. He predicted a breakthrough in robotics within 2 to 3 years and said that integrating AI into physical processes—ordering, follow-up, inspection, transport, sorting, assembly, and quality control—would expand AI's scope from content generation to tangible outcomes. That physical-world integration, he argued, would be a major variable across industries.
Article translated by Jingyue Hsiao and edited by Joseph Chen
