Google I/O Afterparty: Veo, Mariner, and NotebookLM leaders on human-AI collaboration
Recorded shortly after Google I/O, this episode gathers three Google Labs product leads to discuss why public perception of Google in AI has shifted and what comes next. The hosts frame this moment as the end of "chapter one" and the start of "chapter two" for Google, citing strong leaderboard performance and rapid model improvements finally meeting a broad product rollout. All three guests describe years of internal groundwork that is now surfacing. Thomas Iljic (Whisk and Flow) argues that creation must be iterative and "show and tell" rather than long text prompts. He predicts the line between video generation, simulation, and games will blur into world-building experiences where users can pause, reshoot, and regenerate — eventually sharing scenes others can step into. He sees Veo 3's co-generated audio as a major unlock and expects the next 12 months to be defined by remixable video content. Jaclyn Konzelmann (Project Mariner) details the evolution of Google's computer-use agent from a foreground Chrome extension to a background virtual-machine runner capable of 10 parallel tasks. She frames the future in three layers — smarter agents, omnipresent environments (local device, VM, other devices), and agent-to-agent ecosystems — and believes agents will dismantle e-commerce friction, potentially reshaping business models since agents won't be distracted by ads. Simon Tokumine (NotebookLM) explains how Audio Overviews became an accidental viral hit, validates the "audience of one" thesis for personalized media, and previews more show types, mobile-native features, and longer-running project workflows.
Original sources
Podcast Training Data
Episode:Google I/O Afterparty: The Future of Human-AI Collaboration, From Veo to Mariner ↗
YouTube Training Data
Video:Google I/O Afterparty: The Future of Human-AI Collaboration, From Veo to Mariner ↗
AI Radar wrote this summary from the publisher’s notes with AI, not from a transcript. How it works.