AI Mania

AI mania is a sociotechnical condition in which collective belief in the transformative power of artificial intelligence becomes so dominant inside an organization that rational evaluation of AI projects is suppressed. The term, popularized by Mitchell Hashimoto in 2026, captures a state in which continued employment and advancement require repeated public profession of faith in AI’s importance, regardless of whether the underlying projects are succeeding. It is distinct from generic AI hype: hype is external and directed at customers and the press, while mania is internal and reshapes career incentives, decision-making, and what can be said aloud inside the organization.

The mechanism that produces mania is a combination of executive signaling, social pressure, and selection effects. When senior leaders publicly commit to AI-led strategies, employees learn that visible alignment with those strategies is rewarded and visible skepticism is penalized. Workers who deliver results without using AI are at risk of being seen as out of step, and in some observed cases have been terminated in favor of less productive but more publicly AI-aligned peers. Over time, those who remain inside the organization are filtered for their willingness to affirm the prevailing view, and internal data about project outcomes becomes unreliable because honest reporting is career-limiting. The dynamic produces a closed information environment in which managers above a certain level cannot learn whether their AI investments are working.

The tradeoffs are asymmetric. Companies that adopt AI too cautiously risk being outcompeted by faster-moving rivals, while companies that adopt it under mania risk redirecting talent, capital, and attention toward projects that produce no measurable return. The most acute cost is internal: the suppression of honest evaluation is itself a leading indicator of organizational decline, because it removes the feedback loop that would otherwise correct misallocation. Mania is also self-reinforcing at the individual level: an executive who has publicly staked their reputation on an AI strategy has strong incentives to suppress evidence of failure rather than adjust course, even when adjustment would be in the firm’s interest.

The open questions are whether AI mania is a transient feature of an early adoption cycle that will correct as measurement matures, or a structural feature of any technology wave powerful enough to reshape labor markets. Historical analogies to dot-com, blockchain, and cloud suggest a mixed record: some sectors experienced a mania phase followed by sharp correction, while others absorbed the technology quietly after the heat dissipated. A related question is whether external pressure — from investors, boards, regulators, or competitive dynamics — can break the internal information suppression that mania requires, or whether the suppression is too well-incentivized to be pierced from outside.