Nothing on my AI team ships without my yes
I have a team of AI agents. I am also the only person in the room. That sounds like a joke until you have to live with it. I work from Manila, alone, and I ask the agents to do specialised jobs the way a small studio would split them. There is a QA seat that checks work before I see the final pile. There is a listing desk for product copy. A publisher's desk that reads a manuscript like an…
I work alone in Manila with a team of AI agents, each assigned a specialized task. The agents draft, check, and file, but they do not publish or create the content. Before anything is posted, I review the material and ask three questions: why I'm writing it, who it is for, and what potential harm it could cause. If I can't answer these, the draft remains in the folder.
Political posts receive an extra 24-hour hold before I review them again. I name my files for easy retrieval and rely on a QA system to flag issues like politics, real faces, unsupported claims, and missing AI labels. I am the final gatekeeper, adding my human touch to ensure accuracy and integrity. The bots generate drafts, but I am the one who signs off on each piece before it goes live.
Written by urgent.news from Dev.to's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.