Silicon Valley is abuzz with the latest revelation from Boris Cherny, creator of Claude Code at Anthropic, in a viral thread on X. Cherny's personal terminal setup has sparked a discussion on the future of software development, with industry insiders calling it a watershed moment for startups.
Cherny's workflow is surprisingly simple yet allows a single human to operate with the output capacity of a small engineering department. He runs five AI agents in parallel, using system notifications to manage simultaneous work streams. While one agent tests code, another refactors legacy modules, and a third drafts documentation.
This approach validates Anthropic's "do more with less" strategy, which emphasizes superior orchestration of existing models over expensive infrastructure build-outs. Cherny exclusively uses Anthropic's heaviest, slowest model, Opus 4.5, despite its higher computational costs, to eliminate the need for human correction time. By maintaining a single file containing AI mistakes and lessons learned, his team creates a self-correcting organism that improves with experience.
Cherny's workflow relies on automation of repetitive tasks, using slash commands and subagents to handle complex operations with a single keystroke. His verification loop, which tests every change made by the AI, is seen as the key to AI-generated code quality, improving results by 2-3x. The reaction to Cherny's thread suggests a shift in how developers think about their craft, treating AI not just as an assistant but as a workforce that can multiply human output by a factor of five.
