Microsoft CEO Satya Nadella stated that corporate competitiveness in the AI era hinges on building a loop where people and AI learn together, not on choosing the right model. In a post on X, he said this AI shift is fundamentally different from past digital innovation, forming a real cognitive loop between people and digital systems. He said companies must grow two kinds of capital simultaneously: human capital (knowledge, judgement, relationships, creativity) and token capital (AI capabilities a company builds and owns). Nadella said, "As AI capabilities grow, the value of human capital does not shrink. It grows even more." He added, "You can hand work to AI. You can even hand over the job itself. But you cannot hand over learning." He stressed that a company’s future depends on how well it builds up learning accumulated by people and AI together. Nadella said companies should convert internal workflows, domain knowledge and accumulated judgement into AI systems, creating an internal evaluation system, a reinforcement learning environment using real work data, and a searchable knowledge base. He warned against a situation where a few AI models monopolise industrial knowledge, comparing it to 1990s outsourcing that led to industrial hollowing. He emphasised a frontier ecosystem where value flows across all companies, industries and countries, stating, "A platform should create more value on top of it than it takes."