Paper frames work agents as model-plus-harness systems for verifiable tasks
Original title: Work Agents: Systems, Capabilities, and Optimization
It defines work persistence and openness, five work forms, and model, harness, and joint optimization.
A cs.AI paper argues that research lacks a unified view of how agents complete delegated digital work. It treats work agents as foundation models plus runtime harnesses operating in stateful workspaces and producing verifiable outcomes. The authors classify five forms of digital work, name five capability families, and review optimization and evaluation by work outcomes.
- Working conditions are characterized by work persistence and work openness.
- Digital work is split into five forms by task scope and organization.
- Future topics include sustained delegation and adaptation to open environments.
From the original article
1 min read at arxiv.orgAs foundation models evolve from information generation to digital work execution, agents are increasingly capable of completing complex tasks across tools, applications, and workflows. However, existing research lacks a unified perspective on how agents reliably accomplish delegated work.
By Zishan Xu · arxiv.org
A short excerpt; the article and its images belong to the publisher. Summary, key facts and assessment above are AI-written from the article and can be wrong.