Particle.news

Survey Finds Deep Divide Over AI Productivity Between Bosses and Workers

New findings highlight a gap between perceived and measured efficiency, with error-checking eroding much of the claimed time saved.

Overview

  • AI consulting firm Section surveyed about 5,000 white-collar employees and found 33% of executives reported saving 4–8 hours weekly and 19% claimed more than 12 hours, while 40% of non-managers said AI saved them no time.
  • Among workers, 27% reported saving less than two hours per week and only 2% reported saving more than 12 hours, underscoring a sharp perception gap with leadership.
  • Workday’s survey, cited by the Wall Street Journal, found that 85% of employees who credited AI with time savings spent that time correcting AI-generated errors.
  • A METR randomized trial reported experienced open-source developers took 19% longer on tasks when using AI tools even as they believed the tools made them about 20% faster.
  • Section also reported low use-case maturity, with 85% at no or beginner-level use and 40% saying they would be fine never using AI, and noted gains concentrate in tech while retail lags.