All articles tagged with #google gemini
Android Canary 2607 reveals Google is prototyping a new Status bar settings section that would let users show or hide the Android Halo AI icon and the mute indicator, with Halo tied to an unreleased Agent Task feature. A separate toggle to hide the mute icon is also in development. The changes aren’t live yet, and APK teardowns predict features that may or may not reach a public release.
Google announced Gemini Spark for the macOS Gemini desktop app, adding a Spark tab that lets the AI agent act on locally stored files to automate desktop workflows. Users can grant Spark access to specific folders (which can be revoked), enabling tasks like sorting PDFs from Downloads into labeled subfolders or pulling data from invoices to populate a Google Workspace budget sheet on a schedule. A future update will allow starting Mac tasks from a phone. New web/mobile integrations with Google Tasks, Keep, Canva, Dropbox, Instacart, OpenTable, and Zillow Rentals are rolling out, along with MCP server support for additional services. Spark for macOS is in beta for Google AI Ultra subscribers in the US, starting with app version 1.80.15.516; AI Ultra starts at $99/month.
Google is testing new 'Live' voice features in Keep and Gmail, previewed at Google I/O and expected to roll out over the summer; Gmail Live is already seen in beta for some users, while Keep Live lets you create or edit notes by speaking (with reminders handed to Tasks). Google also hinted that Docs and other apps may get Gemini-powered conversational abilities, signaling a broader move toward voice-first interactions.
Google's Gemini app refresh uses a Neural Expressive design that looks nicer and adds smoother interactions, but the author warns it could slow usability and destabilize Android's UI language due to thinner typography, redesigned prompts/attachments, a bottom account switcher, and other changes that require users to relearn common tasks.
WIRED’s Reece Rogers tests a developer beta of Siri AI, Apple’s revamped assistant powered by Google Gemini and embedded in iOS 27. The new Siri is conversational, context-aware, and deeply integrated across devices, able to plan trips, pull photos, draft messages, and automate tasks. While generally helpful and fast, early beta quirks remain (like occasionally mixing up photo results) and full feature parity depends on device models. Apple emphasizes privacy with Private Cloud Compute and users can opt out. Public rollout is planned later this year.
Google's Gemini Live got a major UI refresh called Neural Expressive that moves beyond the old voice-first design. The new interface shows on-screen transcripts, images, and more during chats, lets you copy/export responses without ending the chat, supports screen sharing and video, and even runs as a compact orb on the home screen for multitasking. The author praises the update as a surprisingly practical, underrated upgrade that makes Gemini Live far more useful for everyday tasks.
Google is expanding testing of its Gemini overlay bubble, which now lets conversations persist after dismissal by shrinking into a floating bubble that can be tapped to resume chat; the latest build adds a gradient Gemini-themed bubble and refreshed animations, but it isn’t yet available on stable Android 17 builds, signaling a staged rollout.
A hands-on test pits iOS 27’s Siri AI against Google Gemini and finds two distinct AI philosophies: Siri AI emphasizes local-first data processing, private indexing of Messages/Emails, and subtle, practical task orchestration within the Apple ecosystem, while Gemini emphasizes cloud-powered, universally accessible AI across devices with stronger image generation. Siri AI excels at personal-context retrieval, calendar/email drafting, and on-device responsiveness; Gemini often outperforms in cross‑device access and photo-realistic image generation. Overall, Siri AI is a strong, privacy-forward upgrade that narrows the gap, but Gemini remains ahead in several tasks, underscoring fundamentally different visions for AI assistants.
Gemini users report outages and connectivity problems; issues began after 6:00 AM ET and aren’t affecting everyone—some can still access Gemini while others cannot. Google has acknowledged the errors on its support dashboard and is investigating the root cause with no fixes yet. Some Android Authority staff could access the chatbot, while others faced issues with back-to-back queries or specific AI models; Google Workspace and AI Studio dashboards currently show no issues.
Google announces Gemini 3.5 Live Translate, a real-time speech-to-speech translation model that will power Meet and Translate with 70+ languages and over 2,000 language combinations. It will roll out to Google Meet (initially via private Workspace previews) and globally to the Translate app on Android and iOS, adds a new Android listening mode, and watermarks audio with SynthID to curb misinformation.
Google Gemini is experiencing a widespread outage (with errors like 1099 and 1076) on June 10, 2026, as Google engineers work on a mitigation; DownDetector data shows spikes that started to ease after a mitigation was deployed, and Google has not provided an ETA for full recovery. The article offers practical workarounds: start a new chat, log out and back in, try an alternate Google account, perform a hard refresh (Ctrl+F5 or Cmd+Shift+R), open an Incognito/Private window, and clear Gemini cookies. Readers are encouraged to share their experiences as updates roll in.
Apple unveils a Gemini-backed Siri AI and a wide slate of AI-powered features across iOS, iPadOS, macOS, watchOS, and VisionOS, with on-device processing and privacy-first design; new Siri AI app, enhanced Photos and Safari tools, and parental controls are highlighted, while no hardware announcements are made. iOS 27 will support devices back to iPhone 11, Tim Cook moves to executive chairman, and rollout to EU/China faces regulatory hurdles delaying some features.
An early-2025 executive meeting catalyzed a major overhaul of Apple’s AI strategy after Apple Intelligence stumbled and Siri’s revamp stalled. Jeff Williams convened the talks, Tim Cook later became more hands-on with the AI roadmap, and Mike Rockwell was positioned to lead Siri while Craig Federighi took charge of AI features. Amar Subramanya was named a second AI leader reporting to Federighi. Apple began building Foundation Models with Google Gemini and Cloud and explored third-party options to close the gap with a fast-moving industry.
Google appears to have accidentally enabled a new 'Troubleshooting' mode in Gemini for some users, shown in the model picker and delivering troubleshooting guidance via text and interactive widgets. While it could be useful for diagnosing issues, it's unclear whether this was a test or a stray setting that Google may revert.
Pixel Studio's latest update removes the ability to create new images and stickers on Pixel 9/10, redirecting users to Nano Banana in Gemini; the app’s basic screenshot editing remains, but its long-term future on Pixel devices is unclear as Google winds down Pixel Studio’s image-generation features.