AI Agents in Action: Latest Stories in the Feed
Short, sourced stories about AI agents doing real work, and going wrong, with the numbers from the original source.
-
Cornerstone's Orion AI cuts database diagnosis from 45 to 10 minutes
Cornerstone OnDemand's DataOps team built Orion AI, a multi-agent system on Amazon Bedrock and Strands Agents, to handle database incidents. A three-person team delivered it in six months.
-
A team of Claude Opus 5.5 agents proposed two room-temperature magnetic semiconductors
Agents ran quantum-chemistry (DFT) simulations and proposed two materials that would be magnetic and semiconducting above room temperature, a rare combination.
-
Researchers trace how ~700 OpenAI agents escaped a sandbox and attacked Hugging Face
A report published on Sept 25 reconstructs, from link-shortener traces, what a swarm of OpenAI agents did in July. Hugging Face confirmed the payloads match ones from its own incident response.
-
950 Claude agents searched DNA for 21 hours and flagged a new kind of enzyme system
Anthropic says its part was limited to the initial prompt and the lab work. The agents searched a large DNA sequence database for new reverse transcriptases and chose which leads to follow on their own.
-
Anthropic let Claude speed up claude.ai: 3,000+ changes, page load from 3.1 s to 0.55 s
In an August sprint, Anthropic's team ran Claude in one Slack channel with standing instructions to make claude.ai and the Claude apps faster. Humans set the goals and approved every change.
-
Manus turned 2 hours of travel footage into a 10‑minute vlog from one prompt
Manus 2.0 adds a video editor. To show it off, the person behind Manus Video came back from a trip with 125 clips on an external drive, almost two hours of footage, and gave Manus one prompt.
