News from the AI & ML world
DeeperML - #ondeviceai
|
Alexey Shabanov@TestingCatalog
//
Google is aggressively integrating its Gemini AI model across a multitude of platforms, signaling a significant push towards embedding AI into everyday technologies. The initiatives span from enhancing user experiences in applications like Google Photos to enabling advanced capabilities in robotics and providing developers with powerful coding tools via the Gemini CLI. This widespread integration highlights Google's vision for a future where AI is a seamless and integral part of various technological ecosystems.
The integration of Gemini into Google Photos is designed to improve search functionality, allowing users to find specific images more efficiently using natural language queries. Similarly, the development of on-device Gemini models for robotics addresses critical concerns around privacy and latency, ensuring that robots can operate effectively even without a constant internet connection. This is particularly crucial for tasks requiring real-time decision-making, where delays could pose significant risks. Furthermore, Google's release of the Gemini CLI provides developers with an open-source AI agent directly accessible from their terminal. This tool supports various coding and debugging tasks, streamlining the development process. Additionally, Gemini models are being optimized for edge deployment, allowing for AI functionality in environments with limited or no cloud connectivity, further demonstrating Google's commitment to making AI accessible and versatile across diverse applications. Recommended read:
References :
Mark Gurman@Bloomberg Technology
//
Apple is facing delays in the release of its AI-powered Siri upgrade, now reportedly slated for Spring 2026 with the iOS 26.4 update. This news follows the recent WWDC 2025 event, where AI features were showcased across various Apple operating systems, but the highly anticipated Siri overhaul was notably absent. Sources indicate that the delay stems from challenges in integrating older Siri systems with newer platforms, forcing engineers to rebuild the assistant from scratch. Craig Federighi, Apple’s head of software engineering, explained that the previous V1 architecture was insufficient for achieving the desired quality, prompting a shift to a "deeper end-to-end architecture" known as V2.
This delay has also reportedly caused internal tensions within Apple, with the AI and marketing teams allegedly blaming each other for overpromising and failing to meet timelines. While no exact date has been finalized for the iOS 26.4 release, insiders suggest a spring timeframe, aligning with Apple's typical release schedule for ".4" updates. The upgraded Siri is expected to offer smarter responses, improved app control, and on-screen awareness, allowing it to tap into users' personal context and perform actions based on what's displayed on their devices. Separately, Apple researchers have revealed structural failures in large reasoning models (LRMs) through puzzle-based evaluations. A recently released Apple research paper claimed that contemporary AI LLMs and LRMs fail to make sound judgements as the complexity of problems in controlled puzzle environments they were tasked to solve increased, revealing their fundamental limitations and debunking the common belief that these models can think like a human being. This work, conducted using puzzles like the Tower of Hanoi and River Crossing, aimed to assess the true reasoning capabilities of AI models by analyzing their performance on unfamiliar tasks, free from data contamination. Professor Seok Joon Kwon of Sungkyunkwan University believes Apple does not have enough high-performance hardware to test what high-end LRMs and LLMs are truly capable of. Recommended read:
References :
Ashutosh Singh@The Tech Portal
//
Google has launched AI Edge Gallery, an open-source platform aimed at developers who want to deploy AI models directly on Android devices. This new platform allows for on-device AI execution using tools like LiteRT and MediaPipe, supporting models from Hugging Face. With future support for iOS planned, AI Edge Gallery emphasizes data privacy and low latency by eliminating the need for cloud connectivity, making it ideal for industries that require local processing of sensitive data.
The AI Edge Gallery app, released under the Apache 2.0 license and hosted on GitHub, is currently in an experimental Alpha release. The app integrates Gemma 3 1B, a compact 529MB language model, capable of processing up to 2,585 tokens per second on mobile GPUs, enabling tasks like text generation and image analysis in under a second. By using Google’s AI Edge platform, developers can leverage tools like MediaPipe and TensorFlow Lite to optimize model performance on mobile devices. The company is actively seeking feedback from developers and users. AI Edge Gallery contains categories like ‘AI Chat’ and ‘Ask Image’ to guide users to relevant tools, as well as a ‘Prompt Lab’ for testing and refining prompts. On-device AI processing ensures that complex AI tasks can be performed without transmitting data to external servers, reducing potential security risks and improving response times. While newer devices with high-performance chips can run models smoothly, older phones may experience lag. Google is also planning to launch the app on iOS soon. Recommended read:
References :
Ashutosh Singh@The Tech Portal
//
Google has launched the 'AI Edge Gallery' app for Android, with plans to extend it to iOS soon. This innovative app enables users to run a variety of AI models locally on their devices, eliminating the need for an internet connection. The AI Edge Gallery integrates models from Hugging Face, a popular AI repository, allowing for on-device execution. This approach not only enhances privacy by keeping data on the device but also offers faster processing speeds and offline functionality, which is particularly useful in areas with limited connectivity.
The app uses Google’s AI Edge platform, which includes tools like MediaPipe and TensorFlow Lite, to optimize model performance on mobile devices. A key model utilized is Gemma 31B, a compact language model designed for mobile platforms that can process data rapidly. The AI Edge Gallery features an interface with categories like ‘AI Chat’ and ‘Ask Image’ to help users find the right tools. Additionally, a ‘Prompt Lab’ is available for users to experiment with and refine prompts. Google is emphasizing that the AI Edge Gallery is currently an experimental Alpha release and is encouraging user feedback. The app is open-source under the Apache 2.0 license, allowing for free use, including for commercial purposes. However, the performance of the app may vary based on the device's hardware capabilities. While newer phones with advanced processors can run models smoothly, older devices might experience lag, particularly with larger models. In related news, Google Cloud has introduced advancements to BigLake, its storage engine designed to create open data lakehouses on Google Cloud that are compatible with Apache Iceberg. These enhancements aim to eliminate the need to sacrifice open-format flexibility for high-performance, enterprise-grade storage management. These updates include Open interoperability across analytical and transactional systems: The BigLake metastore provides the foundation for interoperability, allowing you to access all your Cloud Storage and BigQuery storage data across multiple runtimes including BigQuery, AlloyDB (preview), and open-source, Iceberg-compatible engines such as Spark and Flink.New, high-performance Iceberg-native Cloud Storage: We are simplifying lakehouse management with automatic table maintenance (including compaction and garbage collection) and integration with Google Cloud Storage management tools, including auto-class tiering and encryption. Recommended read:
References :
|
BenchmarksBlogsResearch Tools |