AI Psychosis Reaches the Executive Suite

Key Takeaways

• 90% of surveyed firms report zero AI impact on productivity.
• CEOs average under one hour of AI use per week.
• $690 billion AI infrastructure spend expects $600 billion annual revenue.
• Only 20% of AI projects deliver measurable ROI.
• 95% of enterprise AI pilots fail to scale beyond labs.

Pulse Analysis

The AI boom has driven corporations to allocate unprecedented capital toward infrastructure, yet the NBER study shows that the bulk of this spending is not translating into tangible performance gains. With a $690 billion global build‑out aimed at supporting future AI workloads, investors and boardrooms are betting on a revenue horizon of $600 billion—far beyond the current $50‑100 billion generated. This disparity underscores a classic technology adoption curve where early optimism outpaces practical utility, creating a fiscal drag on balance sheets.

Executive engagement with AI tools remains surprisingly low. CEOs log under an hour per week, and even rank‑and‑file employees average only 1.5 hours. Such limited hands‑on experience fuels what Jake Handy dubs "AI psychosis": a belief that AI will revolutionize business despite scant personal interaction. The cultural gap between boardroom expectations and day‑to‑day usage hampers the identification of genuine use cases, leading to a proliferation of pilots that stall in the lab.

For investors and corporate strategists, the data signals a need for disciplined ROI assessment. With only 20% of AI projects delivering measurable returns and a mere 2% achieving transformational impact, rigorous pilot selection and scaling frameworks become essential. Companies may shift toward incremental, revenue‑adjacent AI applications rather than massive, speculative infrastructure bets, aligning spend with realistic adoption rates and preserving shareholder value.