Factory's $50M bet on "agnostic" AI coding agents arrives as the market consolidates around single vendors. With Nvidia and J.P. Morgan backing their multi-LLM approach, the startup claims enterprise gains that challenge typical adoption curves.
Good Morning from San Francisco, Google flipped the switch on conversational search. Voice and video queries now run standard across
Google transforms search from typing queries to live conversations with AI. The feature exits experimental status and becomes standard across US mobile apps, forcing competitors to match conversational capabilities or risk appearing outdated.
Google transforms search from typing queries to live conversations with AI. The feature exits experimental status and becomes standard across US mobile apps, forcing competitors to match conversational capabilities or risk appearing outdated.
💡 TL;DR - The 30 Seconds Version
📱 Google's Search Live launches nationwide September 24, letting users have real-time voice and video conversations with AI search instead of typing queries.
🔬 The feature runs on a custom Gemini version through Project Astra framework, but maintains search DNA by surfacing web links alongside AI responses.
📊 Google processes over 8 billion searches daily, each now potentially convertible to an AI conversation through the new Live button integration.
🚪 Available through two entry points: Live button in Google app or through Google Lens, with no experimental Labs opt-in required anymore.
🏆 Competitors must now match conversational search capabilities or risk positioning as outdated platforms compared to Google's integrated approach.
🔮 Search behavior shifts from query-response to ongoing conversation, fundamentally changing information discovery patterns across the web.
Google's Search Live officially launched across the United States this week, bringing real-time voice and video conversations to the core search experience. The feature exits Google Labs and becomes standard in the Google app for iOS and Android, requiring no experimental opt-in.
Users can now point their camera at objects, speak questions aloud, and receive immediate AI responses alongside traditional web links. The capability represents Google's most significant integration of multimodal AI into its flagship search product.
Search Live runs on a custom version of Gemini, Google's flagship AI model, but maintains the essential character of web search. Unlike Gemini Live's pure conversational approach, Search Live surfaces relevant web links in real time while users speak, preserving the research-oriented DNA of traditional search.
The technical implementation allows for continuous camera input and voice recognition, processing visual and audio data simultaneously. Google processes this through its Project Astra framework—the same multimodal AI system demonstrated at Google I/O in May.
The feature works through two entry points: a dedicated "Live" button beneath the search bar in the Google app, or through Google Lens with a new live conversation option. Both maintain full-screen interfaces with visual feedback indicating when users speak versus when Google responds.
Google's approach differs markedly from standalone AI assistants. While ChatGPT and Claude operate as separate conversational platforms, Search Live embeds AI capabilities directly into the world's most-used search engine.
This integration strategy creates immediate distribution advantages. Rather than asking users to adopt new apps or change behaviors, Google enhances an existing daily habit. The company processes over 8 billion searches daily—each now potentially convertible to an AI conversation.
The timing aligns with broader industry moves toward multimodal AI. OpenAI's Advanced Voice Mode gained traction through 2024, while Anthropic expanded Claude's visual capabilities. Google's response integrates these features into its most valuable real estate: the search bar.
Early demonstrations focus on hands-free assistance scenarios. Travel planning while packing, troubleshooting electronics setup, learning new skills like matcha preparation—all through continuous voice and visual interaction.
The troubleshooting angle appears particularly strategic. Complex technical problems often require back-and-forth clarification that traditional search handles poorly. A user setting up home theater equipment can point their camera at cable configurations and receive immediate guidance, with follow-up questions handled naturally.
Educational applications get significant emphasis in Google's positioning. Students can conduct science experiments while receiving real-time explanations of chemical reactions, or analyze historical artifacts through visual recognition paired with contextual information.
The privacy implications extend beyond traditional search tracking. Search Live captures continuous audio streams during conversations and processes live camera feeds from users' environments.
Google's privacy policy covers the collection but doesn't detail retention periods for conversational audio or visual data processed through the service. The company's standard practice involves storing search queries for personalization—now expanded to include speech patterns and environmental visual context.
This represents a significant expansion of Google's data collection beyond text queries and click behavior. The company gains access to users' voices, immediate surroundings, and real-time problem-solving patterns.
Search Live creates new switching costs for users who integrate the feature into daily workflows. Unlike simple text searches that work identically across platforms, conversational AI searches develop user-specific interaction patterns and response quality.
The feature draws from Google's existing infrastructure—the same knowledge graph, web index, and personalization systems that power standard search. Rival platforms face the challenge of building both conversational AI capabilities and decades of search infrastructure investment simultaneously.
The pattern repeats Google's historical approach to new technologies: absorption rather than separation. Gmail gained smart composition. Maps integrated real-time traffic. Search now incorporates conversational AI. The strategy consistently reinforces existing product usage rather than fragmenting user attention.
The US-only, English-language launch indicates technical caution rather than market confidence. Google's previous AI features—from Bard's early hallucinations to AI Overviews' accuracy problems—demonstrate the risks of global deployment before cultural and linguistic adaptation.
Moving from Labs to general availability suggests Google believes the core functionality works reliably. The geographic restrictions likely reflect regulatory complexity rather than technical limitations. European privacy frameworks require explicit consent for audio and visual processing. Authoritarian markets may impose restrictions on real-time AI capabilities that involve camera access.
International expansion will demand navigation of distinct regulatory environments, cultural contexts, and linguistic variations. The legal compliance often presents greater obstacles than the underlying technology. Each major market will require specific privacy frameworks, data handling protocols, and cultural adaptation of AI responses.
The transformation follows a clear trajectory: search evolves from information retrieval into interactive consultation. Google leverages its distribution advantage while expanding data collection significantly. The integration approach forces competitors to match conversational capabilities or risk positioning as outdated platforms.
• Search behavior shifts from query-response to ongoing conversation, fundamentally changing information discovery patterns and user expectations across the web
• Platform competition now centers on AI integration depth rather than standalone AI products, advantaging companies with existing user bases and infrastructure scale
Q: Is Search Live free to use and what do I need to access it?
A: Yes, Search Live is completely free with no subscription required. You need the Google app on Android or iOS in the US, with no Labs opt-in needed since September 24, 2025. The feature currently works in English only.
Q: What exactly does Google store from my voice and video conversations?
A: Google's privacy policy covers audio and visual data collection but doesn't specify retention periods. The company typically stores search queries for personalization, now expanded to include speech patterns and environmental visual context from camera feeds during conversations.
Q: What are the current limitations of Search Live?
A: Search Live works only in English and is limited to the US market. It can't process multiple languages simultaneously or work offline. Complex research queries requiring multiple sources may still work better with traditional text search and multiple browser tabs.
Q: When will Search Live expand to other countries?
A: Google hasn't announced international expansion dates. European markets require different privacy frameworks for audio and visual processing, while some governments may restrict real-time AI with camera access. Each major market needs specific legal compliance work first.
Q: How does this differ from ChatGPT's voice mode or other AI assistants?
A: Search Live maintains search DNA by showing web links alongside AI responses, unlike pure conversational AI. It integrates directly into Google's existing search infrastructure rather than operating as a separate platform, leveraging the same knowledge graph and web index that power Google's 8 billion daily searches.
Voice typing finally rivals keyboards—if you choose the right balance of privacy, speed, and cost. New AI-powered apps transform rough speech into polished text, but the local vs. cloud decision shapes everything else.
AI music jumped from novelty to seven-figure Spotify plays, but copyright lawsuits now shadow the best tools. This buyer's guide maps which platforms balance quality with legal safety—and why licensing talks may reshape everything.
Privacy-focused email promised liberation from Big Tech surveillance. Reality delivered Bridge daemon crashes, mobile search gaps, and calendar sync headaches. A year later, deadline-driven pragmatism wins over ideological purity.
Claude Code's new sub-agents let you build specialized AI assistants instead of one generalist trying to do everything. Each operates in its own context window, preventing pollution and delivering higher quality results for specific tasks.
Get the 5-minute Silicon Valley AI briefing, every weekday morning — free.
