Summary of Key Points
Recently, major internet companies such as Tencent, JD.com, ByteDance, and Baidu have all made changes to their middle management structures: JD.com has eliminated two levels of management positions, Tencent is piloting a new "responsible person" system that removes the titles of directors and team leaders, ByteDance emphasizes that managers need to "get things done," and Baidu has simplified its job title hierarchy. There are three main reasons for these changes: AI has replaced the core value of middle management in processing information, companies are under significant cost pressure (due to heavy investment in AI), and middle management has become a source of organizational friction (slow decision-making and departmental barriers). However, the article argues that cutting down on middle management is only a superficial solution; it does not address the underlying issues of "peak culture" (conflicts of interest between departments and internal strife). True organizational transformation in the AI era requires redefining the value of roles and breaking the existing structure of power distribution.
Why Are Major Companies Suddenly Making These Changes to Middle Management?
There are three practical reasons behind these changes:
1. AI is taking away middle management's jobs: Middle managers used to rely on information asymmetry—breaking down company strategies into tasks and processing data from lower levels for reports. But AI can now automatically collect data, create presentations, and manage projects, eliminating this advantage.
2. Financial constraints: Companies are spending heavily on AI (for example, Tencent invested 31.9 billion yuan in AI in the first quarter of 2026, which is half of its quarterly profits). To save costs, they need to reduce unnecessary expenses, such as those associated with middle management.
3. Middle management has become a source of friction: In times of growth, additional layers of approval may not be a big problem. But in a period of stagnation, slow product iteration caused by inefficient middle management can be detrimental to a company's success.
What Was the Original Value of Middle Management, and How Does AI Make It Obsolete?
The primary role of middle managers was to act as information intermediaries:
- They informed higher-ups about what was happening at lower levels and helped lower-level employees understand company strategies.
- However, AI can now perform all these tasks efficiently: automatically collecting data, generating reports, and managing projects.
Does Cutting Down on Middle Management Solve the Real Problems?
The article argues that while cutting middle management is a correct approach, it does not solve the root problem of "peak culture"—the conflict of interests between departments. To truly transform an organization in the AI era, it is necessary to redefine the value of roles and change how resources are distributed.
Are the New Systems (Like Tencent's "Responsible Person" System) Reliable?
Tencent's new system transforms management from a permanent position to a temporary one, where someone has authority only while they are in charge. This is similar to the ancient Chinese practice of separating official titles from actual responsibilities. However, this system has potential drawbacks:
- Uncertainty for subordinates: They don't know if the person in charge will still be there next month and whether their decisions will be respected.
- Unfulfilled promises: Promises made by temporary managers may not be kept if they are no longer in position.
How Should Organizations Change in the AI Era?
True transformation requires more than just reducing staff. It involves:
1. Redefining role values: Instead of relying on middle management to process information, companies need to focus on roles that AI cannot perform (such as strategic decision-making and employee support).
2. Breaking the culture of conflict: Changing how resources are allocated so that cross-departmental collaboration is beneficial for all involved.
In short, simply reducing middle management is not enough; organizations need to fundamentally change their approaches to ensure they can adapt to the AI-driven world.