Overview
- Harvard Business Review details an ongoing UC Berkeley ethnography inside a 200-person tech firm over eight months with more than 40 interviews.
- Researchers report the company did not impose higher targets, yet AI made additional tasks feasible that employees voluntarily absorbed.
- Time savings were consumed by expanded to-do lists as work encroached on lunch breaks and evenings.
- Participants described mounting fatigue and professional burnout as expectations for speed and responsiveness grew.
- The findings align with earlier evidence showing modest gains, including an NBER estimate of about 3% time savings and a test where senior developers spent 19% longer despite feeling 20% faster, with forum anecdotes citing tripled expectations for roughly a 10% output lift.