Wire flash
Google unveils persistent Gemini agent as autonomous coworker, leaving humans only judgment
Editorial responsibility
- No named human review is recorded for this page.
- Source reporting is collected, normalized, translated or condensed automatically when needed.
- Automatically published source-backed update
Google has announced a new Gemini agent, described not as a sidebar in Docs but as a persistent coworker with its own identity, memory that survives the device, and the ability to spin up sub-agents. The agent can be handed an outcome and execute it, with skills published by the user's team and models it can swap underneath. The post, attributed to analyst Dion Hinchcliffe, notes that the remaining job for humans is judgment: deciding what to delegate, what context the agent should never have, and what 'finished' means. The announcement was made via a link to Google Cloud, indicating the agent is part of Google's broader cloud and AI strategy.
Source report
Google's latest bet on agentic AI suggests that work begins with a prompt—and often ends long after you close your laptop.
The newly announced Gemini agent is not designed as another sidebar in Google Docs. Instead, it functions as a persistent coworker in the "Claw" style, featuring:
- Its own identity
- Device-independent memory
- Sub-agents it can spin up on demand
- Skills published by your team
- Interchangeable underlying models
Users hand it an outcome, and the agent executes the task.
The Remaining Challenge: Judgment
The critical job that remains is human judgment—deciding:
- What to delegate
- What context the agent should never access
- What "finished" actually means
Good to see @GoogleCloud in this race: https://t.co/F043JZWZzU
Source
dhinchcliffeNeutral / independent
Part of this Story
Google launches persistent Gemini AI agent for enterprise work with subagent delegation