Google just cooked Apple Intelligence

Google's recent keynote unveiled significant smartphone changes, introducing Gemini Intelligence, Android 17, an enhanced Android Auto, and a new laptop category called Google Book. These updates aim to integrate AI deeply into user experience, making Apple software appear less advanced by comparison. Gemini Intelligence introduces five new capabilities. "Rumbler" refines dictated messages, intelligently understanding intent and formatting, even creating lists or translating content. This addresses common dictation frustrations on current devices. Another major improvement is in autofill. Traditional autofill often fails on poorly coded websites, but Gemini Intelligence, accessing a "personal intelligence pocket" of user data like passport details, can smartly fill out any form by understanding the context, not just interpreting code. This promises a much higher success rate and is described as a "seriously fantastic" advancement. Beyond fixing existing pain points, Gemini Intelligence also improves aspects users didn't realize needed attention. Users can now build custom widgets by simply asking for them, making programming accessible even to novices. An example given is a custom marathon countdown, but the potential for location-based activity widgets is highlighted. App automations allow Gemini to process requests across multiple applications. For instance, a user could photograph a tour flyer and ask Gemini to find a similar tour for six people. Gemini would then understand the request, consult an app like Expedia, and notify the user with booking options, without automatically booking to avoid potential user discomfort. Gemini's ability to pick up full screen context enhances its task-doing capabilities. If a user is interested in a standup comedy night, they can ask Gemini to "sort my parking space," and Gemini will autonomously read the necessary information from the screen and make arrangements. This demonstrates Gemini's evolution into an "agent," visibly interacting with apps to complete tasks. This could also extend to creating reminders from a supplement schedule or buying items from a shopping list with a single command. While an infographic summary feature was shown, the result was deemed less impressive, reminiscent of older AI. A significant caveat for Gemini Intelligence is its availability. Features will roll out gradually, starting this summer, primarily on Samsung and Google flagship phones, potentially due to reliance on a more powerful Gemini Nano model. Broader availability could take up to a year. Google is consciously avoiding over-promising, learning from Apple Intelligence's rollout. Android 17, however, is for all users and brings impressive standalone features. It introduces instant "screen reaction" video creation, recording the screen and the user simultaneously, with background removal and proper positioning, streamlining content production. For professional creators, Adobe Premiere is coming to Android with templates for YouTube Shorts, a strategic move to attract influencers and challenge the iPhone's dominance in the creator market. Google is also addressing long-standing quality issues with Instagram on Android. They claim premium Android devices will now produce photos and videos "at least as good, if not better looking" than iPhones. Optimizations to the capture-to-upload pipeline, along with ultra-high dynamic range processing, full night mode, and built-in video stabilization within the Instagram app itself, are promised. Google's direct involvement in fixing this at a base Android level instills more confidence than previous promises from manufacturers. Furthermore, Android 17 allows the phone's native AI to power features within Instagram's editor, such as one-tap photo and video enhancement and audio separation into individual stems. All 4,000 Android emojis are also being remade in 3D. A standout feature in Android 17 is "Pause Point," which is functionally similar to a third-party app called "One Sec." It allows users to set distracting apps, and upon opening them, a 10-second breathing pause is initiated. This pause helps users decide if they truly need the app, redirect to a more useful activity, or proceed with more purposeful usage. While potentially taking market share from the original app, its integration into Android makes it free and more seamless for all users. Unexpectedly, Android Auto emerged as a highlight. All new Android and Gemini Intelligence features extend to the car. Given that drivers are occupied, an intelligent assistant becomes crucial for tasks like ordering food from DoorDash with a voice command. Google Maps will feature immersive 3D navigation, matching real-world views with buildings and overpasses, and improving lane guidance. Cars supporting built-in Gemini will benefit from the AI being trained on the car's hardware, using live camera feeds for precise lane tracking. Gemini will also be aware of the car's dimensions, allowing questions like "Will this TV fit in my trunk?" or "How do I turn on my cruise control?" The new Android Auto will also adapt to various infotainment system shapes, offer updated app aesthetics, and synchronize fonts, wallpapers, and animations with phones. Side widgets for quick contacts or garage door openers are also included. For the first time, Google officially supports 1080p 60fps video playback in cars with Dolby Atmos audio. Videos will minimize to audio-only once driving commences, a clever safety feature, though it requires YouTube Premium for background play. Finally, Google unveiled a new laptop category, the "Google Book," distinct from Chromebooks or Windows. Taglined "intelligence is the new spec," these machines will heavily lean on AI rather than traditional measurable specs. The cursor can be wiggled to activate AI mode, enabling smart features like fusing selected images. Google Books will feature a multicolored light strip to indicate AI activity and will carry over Android features like widget generation. A seamless integration with Android phones allows direct app access and treats phone storage as part of the Google Book's storage, simplifying file transfers. The potential for Google Books to rival MacBook build and trackpad quality could lead users to consider switching from the Apple ecosystem. Overall, Google's new announcements, while not all immediately available, demonstrate a clear focus on reducing "boring admin" through deeply integrated AI, promising a more intuitive and efficient user experience across devices.