Stanford researchers have surveyed 1,500 workers to better understand how employees want AI to handle workplace tasks, revealing striking mismatches between what the tech industry is building and what workers actually want automated.
The study found that while 41% of YC startups focus on areas employees consider low priority, workers mainly want AI to handle repetitive, low-value tasks such as scheduling and data entry to free up time for more important work.
One key contribution of the research is the development of a 'Human Agency Scale' which shows that nearly half of all occupations prefer a balanced human-AI partnership instead of full automation. Arts and media professionals, in particular, show strong resistance to AI replacing their creative tasks, with only 17% of such tasks receiving positive automation ratings.
These insights indicate that while AI agents are likely to become deeply embedded in the future of work, achieving alignment with human preferences will be crucial. Productivity gains could be maximized by focusing on collaboration, not replacement.
To access the full survey and methodology, refer to the paper here.