Nature: Human Scientists Significantly Outperform Best AI Agents on Complex Research Tasks
Nature·medium signal
A Nature study finds human scientists still substantially outperform the best AI agents on complex, multi-step research tasks requiring sustained reasoning and domain expertise. The results challenge narratives about imminent AI-driven scientific automation, suggesting the bottleneck is not raw capability but the ability to maintain coherent research programs across long time horizons. For builders, this frames the current sweet spot as augmentation rather than replacement for knowledge-intensive workflows.