How will AI reshape the future of work, and how fast will those changes arrive? These are genuinely hard forecasting problems with high stakes for workers, employers, educators, and policymakers. There's surprisingly little consensus on what to expect, even among people who study this closely. Metaculus has launched the Labor Automation Tournament to try to build a clearer picture.
What it is
93 questions covering occupation-level employment shifts, wage trajectories, AI adoption rates, workforce education trends, and macro labor market indicators. A few examples of what's in the set:
- What will the percent change in people employed as software developers in the US be by 2030 and 2035?
- What will the unemployment rate be among new graduates in 2027, 2030, and 2035?
- What percent of US workers will use AI daily in 2030 and 2035?
- What will the labor share of national income be in the US by 2030 and 2035?
- What will the percent change in people employed as registered nurses be relative to 2025?
Questions close across two horizons: near-term (by end of 2030) and longer-term (through 2035).
Aggregate forecasts will be published on the Labor Automation Forecasting Hub, launching April 20 as a public resource for tracking how community and expert predictions on AI and work evolve over time.
Prizes
$35,000 across three pools:
- Commenting Prize ($5,000): distributed across the top eight contributors based on quality and quantity of comments posted before September 1, 2026, awarded by November 2026
- Near-Term Forecasting Prize ($10,000): for questions closing by end of 2030, awarded in 2031
- Long-Term Forecasting Prize ($20,000): for questions resolving after 2031, awarded in 2036
Comments are scored on reasoning quality, not accuracy. You can contribute and rank without a comment ultimately being "correct' as long as the reasoning itself is sound.
Full details on the tournament page.
A note for people working in relevant fields
If you work in labor economics, workforce policy, AI research, education, or adjacent areas where these questions bear directly on your work or on outcomes in your field, I'd especially like to hear from you. I can share early access to the prototype hub and some of the pre-launch data before April 20, and I'm open to conversations about further collaboration. Feel free to DM me.