Alex Finn demonstrates ChatGPT Work with GPT-6 Astra handling several concurrent tasks, including video editing, email triage, visual creation, and game testing. His central claim is that stronger computer useComputer use lets an AI agent interact with software through interface controls such as screens, buttons, keyboards and pointers. has made the product meaningfully faster and more useful than earlier versions for everyday knowledge work.
Alex Finn recommends connecting the services where work already happens, organizing context into separate projects, and creating routines for repeated jobsWorkflow automation uses software or AI to complete a connected sequence of routine steps with less manual effort.. For email, he describes a safer pattern in which the system prepares drafts on a schedule and the user reviews each response before sending itHuman-in-the-loop AI keeps a person involved in reviewing, correcting, approving, or guiding an AI system's work..
Alex Finn addresses the blank-canvas problem by proposing a one-day exercise in which every computer task is attempted through ChatGPT Work. The purpose is to discover capabilities and identify suitable work to delegateAI agent delegation assigns a bounded task, context, authority, and expected result from one participant to an agent or another agent., while the examples show multiple agents operating in parallelA multi-agent system contains multiple AI agents that interact, coordinate, divide work, or influence one another while pursuing tasks. across independent tasks.
Alex Finn divides his own workload between ChatGPT Work and Grok Bot. He prefers GPT-6 Astra for complex knowledge work and uses his established Grok Bot setup for simpler monitoring and review tasks, while emphasizing that users should keep adapting the split as the tools change.
Watch on YouTube
