Microsoft CEO Doesn’t Believe AI Will Replace Human Labor: Even If Token Capital Grows, the Value of Human Capital Won’t Decrease

Microsoft CEO Nadella publishes a lengthy article exploring the impact of AI. He introduces the concepts of human capital and token capital, emphasizing that future competition will revolve around building proprietary learning systems and advocating for an open ecosystem.

Satya Nadella introduces the concept of "Token Capital," redefining the core assets of a company

Microsoft CEO Satya Nadella recently (6/14) published a lengthy article discussing the profound effects AI will have on organizational structures.

He believes that the biggest difference between this wave of AI and past digital transformations is that, for the first time, companies can establish a continuous interactive cognitive cycle between humans and digital systems, thereby changing the ways knowledge is accumulated, decisions are made, and intellectual property is created.

Nadella proposes two major concepts: "Human Capital" and "Token Capital."

• The former includes employees' knowledge, experience, judgment, social networks, and creativity;
• The latter refers to the capability assets that companies build and control through AI systems.

He argues that the importance of human capital has not diminished with AI development. On the contrary, AI’s ongoing evolution largely depends on humans setting goals, establishing connections, recognizing patterns, and providing decision guidance. The professional knowledge and experience accumulated by organizations remain a vital source for driving AI capabilities forward.

The key to corporate competition shifts to the "learning cycle," not just a single model capability

Nadella points out that future corporate competition will gradually shift from chasing the most powerful AI models to building their own dedicated learning systems.

He believes that companies can delegate many workflows to AI, but the organization’s own learning ability cannot be outsourced. The knowledge, decision logic, and industry experience accumulated over years need to be transformed into sustainable AI capabilities to create a true competitive advantage.

Within this framework, even if a company replaces its underlying models in the future, it can retain its accumulated knowledge assets. Nadella views this capability as a key sovereignty indicator in the AI era. He believes companies should be able to replace general-purpose models while preserving the professional skills and experience within their organization.

To achieve this, companies need to establish proprietary evaluation mechanisms, internal reinforcement learning environments, and knowledge management systems, enabling AI to continuously learn from actual organizational operations rather than relying solely on public data or generic benchmarks.

AI learning systems are becoming the new intellectual property of organizations

Nadella describes this continuously growing knowledge system as a machine capable of constantly accumulating value. Every process optimization, decision improvement, and work output can generate new training data, further enhancing AI system capabilities.

He believes that future organizational assets will no longer be limited to patents, code, or databases, but will be a learning cycle co-created by humans and AI. These systems will continuously absorb implicit organizational knowledge and gradually form a competitive barrier that is difficult to replicate.

As organizations accumulate more work records, professional judgments, and industry experience, their related AI systems will grow in tandem. Even if the market introduces newer, more powerful large models, latecomers will find it difficult to quickly rebuild the knowledge capital that a company has accumulated over many years.

Warning about over-concentration of AI value and calling for an open ecosystem

Nadella also issues a warning about the future development of the AI industry.

He states that if all the value created by companies ultimately flows to a few large AI model providers, it will impact industry structure and the overall economy.

He cites the offshoring of industries during globalization as an example, noting that while the overall economic scale continues to grow, core capabilities in certain industries and regions are gradually being hollowed out, with ongoing effects. If similar situations occur in the AI industry, the knowledge and expertise accumulated by companies over years could be rapidly commoditized, further weakening their competitive edge.

Therefore, Nadella advocates that industries should prioritize building comprehensive AI ecosystems rather than concentrating all value in a few model platforms. He believes every company, industry, and country should develop their own knowledge cycles and AI capability systems, enabling organizations to continuously accumulate human and token capital and keep value within their own structures and industries.