Theo Browne describes an intensive period using coding agents and compares his reported usage against subscription limits. He presents this as a short-term experiment rather than a sustainable inference budgetAI inference economics examines the revenue, cost, utilization, pricing, and margins involved in serving trained models..
He demonstrates using parallel agentsA multi-agent system contains multiple AI agents that interact, coordinate, divide work, or influence one another while pursuing tasks. to compare pull requests, rank repository work and exchange review notesCode review examines proposed source changes to assess their correctness, clarity and suitability before or around integration.. He also describes accessing a remote Mac for longer jobs and experimenting with agents that check each others output.
Browne acknowledges that producing more code does not establish that it is correct or worth shippingEvaluation is the systematic assessment of a system's behaviour or results against explicit criteria.. The video presents his workflows and model preferences as his experience, with sponsorship omitted.
Watch on YouTube
