Import AI 441: My agents are working. Are yours?
Jack Clark says his research agents processed thousands of papers while he hiked or slept, and Claude finished site scraping, embeddings, local vector search, and a GUI in under one hour. The post confirms multi-agent retrieval, cross-checking, and report generation; it does not disclose model versions, cost, failure rate, or benchmark data. The point to watch is workflow friction dropping enough for AI to shift from single prompts to ongoing delegated work.
Why it matters: A concrete account of research agents running unattended, with the time saved and the tooling built around them.