AI Mania Distorting Corporate Decision-Making, Says Consultant

Nik Suresh, a consultant working with large companies, has documented a pattern of AI-driven dysfunction that goes beyond typical tech hype. In one striking case, an executive confessed to Suresh that they had never used ChatGPT or any AI tool in their life—yet had just produced a technical strategy for an organization with $2B+ in revenue that was entirely centered around AI. The disconnection between experience and strategic authority suggests that AI has become so culturally dominant that executives feel compelled to lead on it without foundational knowledge.

Suresh also shares a report from an engineer at a company with a token leaderboard (a ranking of employees by their usage of AI tools). The engineer describes checking out a parallel copy of their Go repository and telling an AI to rewrite the entire codebase in Zig—a different programming language—while the engineer worked on something else, purely to maintain job security. The motivation is transparent: visibility of AI use has become a job-retention metric, regardless of technical merit.

The most revealing anecdote comes from a conversation between Suresh and a skeptical executive at an enthusiastic company. When Suresh asked why obviously questionable claims were being repeated without opposition, the executive explained that the constraint on honesty was not primarily sales pressure, but rather a structural trap: Customer executives were publicly claiming absurd productivity gains (100x improvements), and if any vendor executive dared to say those gains were not plausible, it would undermine the credibility of the customer executive. Such a challenge would be perceived as an attack or even heresy, and could result in the loss of an enterprise contract. For a vendor executive, getting a contract cancelled over a technical opinion would be career-ending. The result is enforced silence, where the people best positioned to correct inflated assumptions face direct financial penalties for doing so.

The article presents a narrow but revealing snapshot of dysfunction in how large organizations are approaching AI strategy. The core mechanism Suresh identifies is not ignorance alone, but misaligned incentives that systematically suppress honest technical communication. An executive with no hands-on AI experience can produce an AI-centric strategy because there is no enforced accountability for technical credibility; an engineer will undertake wasteful rewrites because career survival depends on appearing AI-productive. Most critically, the vendor-customer dynamic creates a perverse incentive structure: vendor executives know that customer executives have made public commitments to implausible productivity gains (100x improvements), so questioning those gains would create a reputational liability for the customer. The fear of contract loss outweighs the responsibility to offer honest technical counsel. This describes a classic information cascade where public commitment to an assumption becomes self-reinforcing, particularly when the people who could correct it face personal financial consequences for doing so.