AI fails to pay off because it is deployed inside silos
Hastings (University of Washington), Syverson (Chicago Booth) and Zimmerman (Yale SOM) pull together prior evidence. An NBER survey of about 6,000 executives found roughly nine in ten saw no measurable productivity effect from AI, and McKinsey found only 39% of AI-using firms report profit impact. Their diagnosis is that tools are tuned to tasks within one function, while the gains sit in the handoffs between functions.
The scarce, valuable work is redesigning the handoffs between teams, not making one task faster. Knowing how the teams upstream and downstream of you work is what lets you turn a local pilot into a change that pays off for the business. This is a synthesis of existing evidence rather than a new study.
Save this story, pin it to the top of your feed, or keep a private note on it.
Sign in