OpenPrediction
Within a decade, AI agents will independently do many software tasks that take humans days or weeks
“Extrapolating this trend predicts that, in under a decade, we will see AI agents that can independently complete a large fraction of software tasks that currently take humans days or weeks.”
Context: METR, "Measuring AI Ability to Complete Long Tasks" (Mar 19, 2025): 50% time horizon doubling about every 7 months.
How we judge it
Our interpretation: by 19 Mar 2035, METR (or equivalent) reports frontier agents with a 50% time horizon of at least one work-week (40 hours) on software tasks.
Resolution criteria are our interpretation of the claim, written before grading. Methodology
- Said
- Resolve by
- Due in 3084 days
- Verdict
- Open
- Stated probability
- None stated
- Specificity
- 2 of 3 (dated, some interpretation)
