A naukri.com initiative
Analyticsindiamag
3d
116
Image Credit: Analyticsindiamag
Google Gemma 3n is What Apple Intelligence Wants to Be
- Apple introduces 'Foundation Models' framework with a 3-billion parameter on-device model optimized for performance on Apple Silicon.
- The model supports 15 languages, text and image processing, and tasks like summarizing, extracting key information, and original content creation.
- Google's Gemma 3n poses direct competition, operating on phones, tablets, and laptops with text, image, and audio inputs.
- Gemma 3n is praised for its multimodal capabilities, open-source approach, and collaboration with hardware manufacturers for optimized performance.
- Apple Intelligence uses a hybrid approach combining on-device processing with cloud-based computation, unlike Gemma 3n's on-device capabilities.
- Google is integrating AI deeper into the Android ecosystem, allowing smartphone users to run large language models locally through the Edge Gallery app.
- Apple enhances its AI features with Live Translation, Image Playground, Genmoji, Visual Intelligence, and Workout Buddy, integrated deeply into the operating system.
- Apple claims superiority over Qwen and Gemma models in various languages, though Siri received minimal focus during WWDC 2025 due to quality concerns.
- Apple's software chief revealed that Siri demos met internal standards poorly, leading to non-release; Google's pace-setting in AI poses a challenge for Apple's future approach.
- Apple's subtle integration of AI may need to evolve in the face of Google's advancements.
- Google Gemma 3n article is eligible for a web story.
Read Full Article
7 Likes
For uninterrupted reading, download the app