The one number that could change how we talk about AI and work
The recent MIT Technology Review piece highlights a powerful and practical idea: rather than arguing endlessly about hypotheticals, we can collect one clear piece of data that shows how AI is changing jobs in practice. That data is granular, task‑level time‑use information — essentially, reliable measures of how workers spend their hours before and after AI tools are introduced.
Why this matters: Time‑use data lets employers and researchers see which tasks are being automated, which tasks grow, and whether workers are freeing up time for higher‑value activities. With that evidence in hand, companies can design targeted upskilling programs, unions and policymakers can craft better safety nets, and public debate can move from fear to solutions.
Collecting this data responsibly — de‑identified, standardized, and aggregated — creates a win for everyone. It improves transparency, helps measure true productivity gains, and highlights where job transitions are occurring so training resources can be directed where they’ll do the most good.
Practical benefits:
- Employers can quantify real productivity improvements and redeploy staff to higher‑value work.
- Policymakers gain evidence to design retraining, wage supports, and labor market interventions that match observed impacts.
- Workers and advocates get a fact‑based foundation for negotiations and career planning instead of speculation.
In short, better measurement is an achievable, positive step that turns anxiety about AI into targeted action. The spotlight on a single, well‑chosen metric could speed fairer adaptation and help ensure AI benefits more people.