Google expands Personal Intelligence / Gemini to consumers
Key Questions
What is Gemini Spark and who can access it?
Gemini Spark is Google's 24/7 proactive personal AI agent now in beta for Google AI Ultra subscribers, handling task execution and background monitoring. It is rolling out to macOS and other platforms with integrations like Canva and Instacart.
How can users opt out of Google using their data to train AI?
A recent privacy settings change allows Google to store more user data for AI training. Users can adjust these settings in their Google account privacy options to limit data sharing.
What new image editing features are available in Gemini for Chrome?
Gemini in Chrome now lets users right-click and edit any web image using the Nano Banana model directly in the browser on Android devices.
What improvements does Gemini 3.5 Live Translate offer?
Gemini 3.5 Live Translate provides real-time speech-to-speech translation in over 70 languages while preserving the original voice, rolling out to Translate and Meet apps.
How does Gemini's memory feature work across platforms?
The memory feature remembers user names, preferences, and context across Google services to deliver more personalized responses and proactive assistance.
Post-I/O: Gemini Spark rolling out to Ultra subscribers (now in beta for task execution), Gemini Omni video editing in India, Google Flow Beta, Dreambeans, Gemini Go, Gemma 4 12B. New: Google World Cup AI tools (live scores, AI Mode, Gemini briefings). Also: Gemini Omni cloning demo shows practical avatar creation. Summer rollouts continue. New: Google AI Plus subscription price drop to $4.99/month for 400GB — competitive pressure on Microsoft Copilot. New: Gemini 3.5 Live Translate rolls out — real-time speech-to-speech in 70+ languages with voice preservation, rolling to Translate and Meet. Google AI Studio hits 18M apps created, 1.2M/week, with native Android and Gmail/Docs integrations. New: Gemini memory feature rolls out — remembers names, preferences across platforms. Firefly help page reveals new model support (Gemini 3, GPT Image 1.5) and Google Search integration. New: Google I/O 2026 reveals Gemini Spark as 24/7 proactive personal AI agent, Gemini 3.5 Flash 4x speed and half cost, $180B capex signals AI infrastructure dominance. New: Gemini Omni Flash rolling out Veo 3.1 video generation to all AI subscribers — major expansion of consumer video creation. New: Gemini Omni Flash tops Video Arena benchmark with big lead; API access rolling out to developers. New: Gemini on Sony Google TVs expands living room presence. New: Google rolls out Information Agents to Ultra subscribers — proactive search with background monitoring, link-rich updates. Shifts search from pull to push. New: NotebookLM Audio Overview gets major customization (formats, length, host guidance). New: June Pixel Drop brings text-to-video via Gemini Omni to Pixel creators. New: Android 17 arrives on Pixel with floating apps, foldable features, and new AI tools: Gemini Omni video gen and Lyria 3 music creation — AI delivered via services, not OS updates. New: Gmail AI summaries roll out to 1B+ users globally. New: Google Finance exits beta with portfolios, AI tools, dedicated Android app. New: Gemini Spark rolling out to macOS app for local task automation (inbox summaries, event planning). New: Google Nano Banana 2 Lite — faster, cheaper AI image generator, $0.034/1k images, 4-second generation, trade-offs in text rendering. Hands-on review highlights speed as game-changer for iteration. (July 21) New: Gemini Spark lands on Mac with file management, topic tracking, and integrations like Canva, Instacart, OpenTable. MCP support for custom apps. Only for Ultra subscribers in US. (July 21) New: Gemini Spark gets further updates (July 21) — ongoing rollout of 24/7 personal AI agent features. New: Hands-on review of Gemini Omni confirms conversational multi-input prompts, available on any phone with subscription, replaces Veo in Gemini app. (July 21) New: Gemini personalized image generation now free for US users — leverages Personal Intelligence data (Gmail, Photos) for tailored outputs. Drives adoption, consolidates Pixel Studio shutdown. Privacy opt-in. (July 21) New: Nano Banana 2 image model now free and default in Gemini app, Photos, Search, NotebookLM across 100+ countries. Fast (~4 sec), face-consistent editing. Tutorial available. (July 22) New: YouTube Create adds Gemini Omni for 10-sec AI video gen and Nano Banana for AI image gen directly in editor — competes with CapCut/Runway for mobile creators. Also adds Edit with AI auto-editing. Limited countries. (July 22) New: Gemini in Chrome now supports editing web images via right-click using Nano Banana — practical integration tip for Chrome users. (July 22) New: Android Auto update with Gemini AI — Magic Cue contextual AI, edge-to-edge Maps, video playback, Dolby Atmos, DoorDash ordering. Millions of vehicles. (July 22) New: Google under-the-radar opt-in for AI training on personal media; users can opt out — important privacy concern for all Google users. (July 22) New: Meta launches Muse Image model — free for consumers, subscription for power users, integrated into WhatsApp/Instagram, personalization via uploaded photos. Benchmarks beat Nano Banana 2 but trail GPT Image 2. (July 22) — competitive pressure on Google's image gen.