Remote Labor Index: Measuring AI Automation of Remote Work
2025/10/30 by Mantas Mazeika, Alice Gatti, Mazeika, Mantas +91 · 9 voices · 9 citations
#cs.LG #cs.AI #cs.CL
paper · pdf · doi:10.48550/arxiv.2510.26787
Abstract
AIs have made rapid progress on research-oriented benchmarks of knowledge and reasoning, but it remains unclear how these gains translate into economic value and automation. To measure this, we introduce the Remote Labor Index (RLI), a broadly multi-sector benchmark comprising real-world, economically valuable projects designed to evaluate end-to-end agent performance in practical settings. AI agents perform near the floor on RLI, with the highest-performing agent achieving an automation rate of 2.5%. These results help ground discussions of AI automation in empirical evidence, setting a common basis for tracking AI impacts and enabling stakeholders to proactively navigate AI-driven labor automation.
Citations
Cited by
Discussions
- Heard the one about how dozens of researchers assigned $144,000 worth of real world human tasks to the world’s “smartest” AIs to manage and… They fucked up so monumentally that they were able to gener [bsky, 119 points, 2 comments]
- Remote Labor Index: Measuring AI Automation of Remote Work Plus de 96% d'échec. Des milliards investis et le résultat va vous surprendre. arxiv.org/abs/2510.26787 [bsky, 4 points, 2 comments]
- Remote workers, aka #freelancers, work better than #AI. The Remote Labor Index is a measure of AI automation capacity. AI agents perform at an automation rate of <3%, a stark gap btwn progress on comp [bsky, 3 points, 0 comments]
- AI Pubblicato nuovo studio RLI sull'efficacia in ambito operativi per così dire umani A chi interessa l'intero studio il link arxiv.org/pdf/2510.26787 Per gli altri breve riassunto👇 [bsky, 3 points, 1 comments]
- Remote Labor Index: Measuring AI Automation of Remote Work [hn, 2 points, 0 comments]
- I've been reading this for half an hour & I can't get over it: <2% average success rate across models at doing a narrow subset of business tasks, and an overall cost reduction of 4%? this is what the [bsky, 2 points, 3 comments]
- Your job is safe! For now. Despite hype, AI cannot work alone. New measuring benchmark highlights how valuable human domain knowledge and project management skills are in freelance projects. Several A [bsky, 1 points, 0 comments]
- "AI agents perform near the floor on Remote Labor Index, with the highest-performing agent achieving an automation rate of 2.5%" arxiv.org/abs/2510.26787 [bsky, 0 points, 0 comments]
- Agentic AI is a scam. It’s totally incompetent at human tasks & needs root access to your device like the scam tech caller who defrauded my dementia-addled mother. [bsky, 0 points, 0 comments]
Related