Nadella urges firms to build AI ‘token’ and human capital

Microsoft Chief Executive Satya Nadella wrote this week that companies should develop both proprietary AI systems, which he calls “token capital,” and employee knowledge and judgment, which he calls “human capital,” to ensure learning and value stay within firms.

Nadella defined token capital as a company’s own AI systems and models. He described human capital as employee expertise, pattern recognition and business relationships that shape judgment. He argued that human direction is necessary for AI to produce useful outcomes and warned that without it “you have compute running in circles.”

He outlined a learning loop built on top of base models in which firm-specific data and experience accumulate across people and AI. That loop would capture institutional learning even when individual tasks or jobs are automated. “You can offload a task, or even a job, but you can never offload your learning. The future of the firm is the ability to compound that learning across people and AI,” he wrote.

Nadella urged businesses to create agentic systems that improve with use and to keep control of their intellectual property. He proposed private evaluations and reinforcement learning environments that use a company’s own data to turn workflows and judgment into systems that get better over time. Those systems should let a company change the underlying base model without losing its accumulated expertise. “This is the key ‘test’ of your control and sovereignty in the era ahead,” he added.

He warned against a concentration of value in a small number of dominant models, comparing that outcome to earlier waves of globalization that hollowed industrial sectors. “The last thing any of us want is a world where every company across every sector is ceding value to a few models that eat everything they see,” he wrote, adding that an AI economy where most value accrues to a handful of models would lack political and social acceptance.

Nadella’s remarks come as enterprise spending on AI is growing faster than many companies expected, prompting questions about whether that investment builds owned capability or increases dependence on outside providers. He recommended combining owned technology and employee judgment so organizations retain control of the knowledge that drives their operations.