Addy Osmani spent over 14 years at Google, most recently as a Director at Google Cloud AI, running Cloud AI Developer Experience and Technical Evangelism and helping launch Google's agent developer platform.
How media typically covers Addy Osmani
Based on 15 scored articles
Addy Osmani as author
AI coding assistants are shifting developer identity from hands-on code creation to AI orchestration, creating psychological and career implications for the profession.
“Author of "Addy Osmani on Vibe Coding and AI-Assisted Engineering" in Google”
Effective AI agent specifications require clear high-level vision, manageable context windows, iterative refinement, and task decomposition rather than comprehensive upfront documentation.
“Author of "How to Write a Good Spec for AI Agents"”
Referenced in coverage
“Confirmed Claude Code is instructed not to rebuild specific existing games over IP concerns.”
“Confirmed that Anthropic added guardrails to Claude Code to prevent game recreation and stated the company is iterating on rules to strike the right balance.”
“Confirmed Claude Code was instructed not to rebuild specific existing games due to IP concerns, while clarifying original games remain supported.”
“Published guidance on how to use Claude Opus 5.5 more effectively by reducing over-prompting and leveraging the model's capabilities with less instruction.”
“Co-authored article on Claude Code team's code review process”
“Author of technical guide on Claude Code mods system that allows users to write TypeScript plugins for AI coding tools.”
“Byline author of article about Claude Sonnet 5.5 best practices from Anthropic.”
“Author analyzing task costs and token pricing on Opus 5.5 model.”
“Published the essay that popularized the name of loop engineering for coding agents.”
“Delivered keynote addressing what AI engineers should focus on as code generation cost approaches zero.”
“Former Google engineering leader who discussed agents running inner loops while humans maintain outer loop oversight.”
“Posted on X about agentic code review, noting that coding agents have made the hard part of engineering the review process rather than code writing itself.”
“Byline author discussing loop engineering as the next evolution of working with coding agents, replacing single prompts with recursive systems.”