The Cognitive Crisis at the Top: Why Declining Deep Judgment Among Executives Threatens the AI Transformation of Work

In boardrooms and C-suites worldwide, a quiet erosion is underway. The capacity for deep, sustained reading, rigorous analysis, robust mental modeling, and long-range extrapolation, the foundations of high-quality strategic judgment, is measurably declining among many knowledge workers, managers, C-suite executives, CXOs, and board directors. Driven by digital attention fragmentation and accelerated by generative AI, this trend produces shallower decisions, greater organizational complexity, and heightened risk precisely when AI demands sharper human oversight.

The consequences appear in how leading firms restructure work. Microsoft CEO Satya Nadella described a deliberate experiment at LinkedIn: collapsing product manager, designer, front-end engineer, and back-end engineer roles into "full-stack builders" with expanded scope. "We used to have product managers, we had designers, we had front-end engineers, and then we had back-end engineers and so on," Nadella said on the All-In Podcast at Davos in January 2026. "So what we did is we sort of took those first four roles and combined them. In fact, increased scope and said, they're all full-stack builders." For AI products, the workflow now "starts with evals... eval to science, to infrastructure," with full-stack builders owning the evaluation loop. LinkedIn ended its Associate Product Manager program and launched the Associate Product Builder track, training early-career talent in code, design, and product management at once. Evals move upstream as integral to building.

The logic is sound. AI tools lower coordination costs across boundaries, enabling higher velocity and tighter ownership. Success, however, depends on the cognitive depth of the people involved. Without strong independent judgment, expanded scope risks confident but brittle outcomes, the "mess" critics such as API and AI coach James Higginbotham have seen in similar experiments.

The Measurable Decline

Evidence is robust. Daily reading for pleasure among U.S. adults fell roughly 40 percent over two decades, from about 28 percent in the early 2000s to 16 percent by 2023. College-educated professionals, the main executive pipeline, saw average books read drop from roughly 21 per year to under 15. Neuroscientist Maryanne Wolf describes the "shallowing hypothesis": digital media weakens the neural circuits for deep reading and critical analysis. "We are not only what we read," Wolf has observed; "we are how we read." Skimming and rapid switching erode inference, synthesis, and insight.

Attention data confirm the pattern. Focused time on a single screen task fell from about 2.5 minutes in 2004 to roughly 47 seconds recently. Knowledge workers face interruptions every couple of minutes. Cal Newport notes in Deep Work that "Deep work is exhausting because it pushes you toward the limit of your abilities." Sustained high-intensity effort rarely exceeds four hours even for practiced individuals, and far less under fragmentation. Cognitive capacity has become the binding constraint.

At the C-suite and board level the effects intensify. Pre-reads go unread. Directors rank information quality and flow from management as both a top driver of effectiveness and a frequent weakness. Cognitive overload and shortfalls in handling complex data are documented in board research. AI compounds the risk through offloading: frequent users show weaker independent critical thinking, with experiments revealing high rates of following incorrect AI advice with elevated confidence. The result is cognitive debt. Judgment softens as synthesis migrates into systems leaders do not fully interrogate.

Charlie Munger put it plainly: "In my whole life, I have known no wise people (over a broad subject matter area) who didn't read all the time—none, zero." Warren Buffett attributes much of his judgment to spending five to six hours a day, or about 80 percent of working time, reading and thinking. These outliers show both the value of the practice and how rare it has become.

AI Restructuring Meets Cognitive Fragility

Nadella's full-stack model and the upstream shift of evals adapt rationally to AI. Private evaluation sets are becoming critical intellectual property. Expanding individual scope can reduce friction and accelerate throughput.

Yet when the people defining evaluations and owning outcomes operate with diminished analytical capacity, risks multiply. Craft depth erodes. Healthy tension between functions collapses into unexamined assumptions. Systems may look innovative yet lack coherence or foresight. Critics note the loss of checks that specialization once provided. Some observers cite LinkedIn's own product record as insufficiently innovative to prove the model's success, highlighting the gap between ambition and quality.

These cognitive trends scale upward. Boards and C-suites that lean on summaries, dashboards, and AI insights without protecting time for primary sources are more likely to endorse structural changes on analyst language rather than first-principles understanding. Complexity then grows: more process, tools, and meetings substitute for clarity.

Protecting Judgment as Strategy

A bifurcation is visible. Leaders and organizations that treat deep cognitive capacity as a scarce, trainable asset, protecting time for long-form reading and thinking, requiring primary-source work, designing AI use that forces verification rather than passive acceptance, and measuring decision quality, will pull ahead. Those who assume tools fully compensate for eroded judgment will accumulate debt that surfaces in missteps and mediocrity.

Wolf's research shows the reading brain remains plastic; circuits for deep analysis can be rebuilt. Newport demonstrates that even busy executives can redesign systems to reduce context-switching. Munger and Buffett prove the highest returns still accrue to consistent investment in input and reflection. Nadella correctly sees that AI changes the unit of productive work. The open question is whether the people inside those new structures will retain the depth required to guide them.

In an era when AI generates fluent answers at scale, the scarce capability is the human ability to sit with complexity, interrogate assumptions, model second- and third-order effects, and exercise independent judgment. Evidence shows this capacity is under pressure at every level, including the top. Rebuilding it is a strategic imperative. The firms and boards that succeed will treat deep reading, rigorous analysis, and clear extrapolation as non-negotiable infrastructure for the age of intelligent machines.

[Major General Dr. Dilawar Singh, IAV, is a distinguished strategist having held senior positions in technology, defence, and corporate governance. He serves on global boards and advises on leadership, emerging technologies, and strategic affairs, with a focus on aligning India's interests in the evolving global technological order.]