Google I/O 2026 keynote in 35 minutes
TL;DR
Google I/O 2026 unveiled a comprehensive shift from reactive AI to autonomous agentic systems, introducing Anti-Gravity 2.0, the persistent Gemini Spark agent, and Gemini 3.5 Flash alongside multimodal creation tools and a redesigned search infrastructure that promises to handle complex tasks continuously in the background.
🤖 Autonomous Agents & Anti-Gravity 4 insights
Anti-Gravity 2.0 Desktop Launch
A new standalone desktop application featuring sub-agents, hooks, and asynchronous task management that enables Gemini to build complex projects like operating systems from scratch.
Gemini Spark 24/7 Agents
Personal AI agents running on dedicated Google Cloud VMs that perform long-running background tasks via voice, email, or chat, integrating with third-party tools through MCP.
Persistent Search Agents
Background monitoring agents that continuously scan the web across sites, social media, and forums for specific criteria like apartment listings or product drops, delivering proactive updates.
Daily Brief Agent
An out-of-the-box agent that synthesizes inbox, calendar, and task data into a concise morning digest organized by topic with suggested next actions.
🧠 New Models & Creative Workflows 4 insights
Gemini 3.5 Flash Speed
A frontier-intelligence model co-optimized with Anti-Gravity that delivers outputs 4x faster than competitors while outperforming 3.1 Pro on coding and economically valuable benchmarks.
Gemini Omni Video Generation
A multimodal model combining video, image, and simulation capabilities with intuitive physics understanding, enabling conversational video editing from any input including user selfies.
Voice-First Productivity Tools
Docs Live enables real-time voice-controlled document creation and editing in Workspace, while Ask YouTube offers conversational search that jumps directly to relevant video segments.
Generative UI in Search
Search now dynamically builds custom interactive visualizations and widgets on demand using Anti-Gravity coding, creating tailored educational interfaces for complex queries like astrophysics.
🔍 Search Evolution & Trust 4 insights
Universal Cart Intelligence
A cross-merchant shopping cart that works across Search, YouTube, and Gmail to track price drops, monitor stock, and find deals automatically in the background.
Multimodal Search Box Redesign
The first major upgrade to Search in 25 years supports text, image, file, and video inputs while unifying AI overviews and AI mode into a seamless conversational experience.
Content Credentials Expansion
Synth ID verification expands to Search and Chrome with new partnerships from OpenAI, Cacao, and Leven Labs, enabling users to verify AI-generated content origins and edits.
Neural Expressive Design
A complete Gemini app redesign featuring fluid animations, vibrant colors, and regional dialect options for Gemini Live, rolling out globally across Android, iOS, and web.
Bottom Line
Google is transitioning from passive AI assistance to autonomous agents that perform complex, multi-step tasks continuously in the background, requiring users to adapt to a new paradigm of delegated digital labor and persistent AI access.
More from The Verge
View all
Everybody wants to rule the AI world | The Vergecast
The Elon Musk vs. OpenAI trial reveals a toxic power struggle driven by control battles and self-dealing, with damning text messages and journal entries exposing how personal conflicts between a handful of tech leaders shaped the AI industry's trajectory while highlighting terrifying future legal risks of AI-assisted discovery.
What an AI-designed car looks like | The Vergecast
Automotive journalist Tim Stevens explains how AI is compressing the traditional 5-6 year car design process into potentially 3 years by automating 3D modeling and wind tunnel simulations, while warning that eliminating entry-level creative tasks could break the talent pipeline for future designers.
Elon Musk had a bad week in court | The Vergecast
Elon Musk's testimony in his lawsuit against OpenAI backfired dramatically as he struggled under cross-examination, admitting that his AI company xAI distilled OpenAI's models and conceding he failed to read key contractual documents before contributing $44 million.
Framework is making PCs cool again | The Vergecast
David Pierce revisits the Rabbit R1 AI device, finding unexpected utility in its voice recording features despite earlier failures, before joining The Verge's Liz Loeffler to analyze the OpenAI vs. Elon Musk trial as a legally weak but damaging act of 'lawfare' driven by personal vindictiveness.