Josh Woodward on Google Labs' 0-to-1 AI Product Playbook
Josh Woodward, who leads Google Labs, describes the organization's mandate to build zero-to-one AI products outside Google's traditional product groups while maintaining a close collaboration with Google DeepMind. He outlines Labs' culture of shipping ideas to users within 50–100 days, celebrating milestones like 10,000 weekly active users, and recruiting unconventional talent—including author Steven Johnson, who co-founded NotebookLM. Woodward frames the current moment as one of accelerating progress, citing the 97x cost reduction in text models over the past year as a leading indicator for video economics. A central theme is the evolution of prompting: Woodward argues paragraph-style text prompts are already archaic for end users, with context increasingly delivered through documents, images, voice, and video. He positions Google's products accordingly—NotebookLM (sources-as-context), Mariner (computer-use agent), and Veo 2 (generative video with high-fidelity physics and a collapsing cherry-pick rate). On Mariner, Woodward says early enterprise use cases like call-center co-browsing and sales follow-ups are showing more traction than consumer tasks, and he expects several model revisions over the coming year as X/Y coordinate navigation and human-in-the-loop controls mature. On generative video, Woodward bets that steerability, character/scene consistency, and an emerging "AI camera" concept will reshape entertainment, with content becoming more interactive, remixable, and personalized. He notes Google is hedging between pixel-stream and 3D-native video approaches. Woodward identifies coding, long/infinite context, and "taste" in application-layer design as underhyped, calls the chatbot interface overhyped, and urges builders to think about whether their products amplify or replace human creativity.
Original sources
Podcast Training Data
Episode:Josh Woodward: Google Labs is Rapidly Building AI Products from 0-to-1 ↗
YouTube Training Data
Video:Josh Woodward: Google Labs is Rapidly Building AI Products from 0-to-1 ↗
AI Radar wrote this summary from the publisher’s notes with AI, not from a transcript. How it works.