In the evolving landscape of artificial intelligence (AI) integration within corporate structures, middle management remains a vital component despite predictions of its obsolescence. Kai-Fu Lee, a prominent technology executive, reflects on the persistence and transformation of middle management roles amid technological advancements, drawing on historical and contemporary examples.
In 1958, management scholars speculated that computers would eliminate middle managers by the 1980s, a prediction only partially realized. While computers automated many managerial tasks, they did not replace the managers themselves. Attempts to flatten organizational hierarchies surfaced again in the early 2000s, notably at Google, where founder Larry Page experimented with reducing layers of management. However, the initial enthusiasm gave way to the reinstatement of middle managers within months, as complex operational and human resource challenges emerged. By 2005, when Lee joined Google to lead operations in China, he relied heavily on multilayered management to navigate the demands of a large, multilingual, cross-cultural workforce.
Today, AI technology is again challenging traditional organization charts, with the potential to reshape company hierarchies more fundamentally. Organizations function as information systems where middle management historically served as a necessary layer to manage communication and data flow. Executives cannot personally absorb every document or maintain direct contact with all employees, so intermediate managers have traditionally filtered and relayed information. AI’s ability to access and analyze vast quantities of company data, such as meeting transcripts, support tickets, and shared documents, improves transparency by providing direct insights to top executives. This capability helps reduce signal degradation that typically occurs across communication layers.
Lee highlights his personal experience using what he terms “Kai-Fu’s Boss AI,” a system that integrates with all corporate records and communications. This AI recently identified two teams unknowingly duplicating work over several months, exemplifying how AI can expose inefficiencies previously overlooked. As the cost and complexity of information reconciliation decrease, many conventional middle management functions may become redundant.
Consequently, some companies are moving toward flatter organizational structures centered on an executive team overseeing directly responsible individuals (DRIs), a concept popularized by Apple. DRIs assume full accountability for specific outcomes, such as products or market segments. In larger organizations, adaptable middle managers may transition into these outcome-focused roles, leveraging AI tools to enhance their effectiveness.
Despite the automation of routine tasks by AI, Lee emphasizes that the core value of human managers will shift rather than vanish. Managers will increasingly focus on exercising judgment, mentoring staff, and handling complex interpersonal dynamics. The experience and institutional knowledge held by seasoned managers remain crucial, especially for guiding teams through uncertainty and facilitating collaboration across departments and external partners.
Lessons from Google’s earlier flattening efforts underscore the risks of rapid, extensive reductions in management layers, which can lead to coordination breakdowns, employee dissatisfaction, and executive burnout. Human relationships and trust-building—skills beyond AI’s current capabilities—continue to undergird organizational success.
As AI assumes greater roles in operational execution, human managers are poised to become more indispensable in fostering collaboration, navigating ambiguity, and sustaining organizational culture, marking an evolution rather than an eclipse of middle management’s role.