A naukri.com initiative
Google News
Siliconangle
1M
0
Image Credit: Siliconangle
Google expands availability of Deep Research, an AI tool that wades deep into any topic
- Google has expanded the availability of its experimental AI research tool, Deep Research, to Gemini Advanced users on Google Workspace.
- Deep Research can efficiently explore complex topics and provide detailed reports by analyzing numerous sources in a matter of minutes.
- It is beneficial for industry research, competitive analysis, customer research, and educational purposes like lesson planning and grant writing.
- Exclusive to paid Google Workspace users with access to Gemini Advanced, Deep Research enables deep research by switching to the '1.5 Pro with Deep Research' model.
- Users can input prompts about topics they wish to research deeply, and Deep Research will generate a comprehensive research plan, refining its analysis as it progresses.
- Once completed, Deep Research provides a detailed report with key findings, original source links, and the ability for users to ask follow-up questions or conduct deeper research.
- Google emphasizes that Deep Research saves time by condensing hours of research into minutes on various topics and will be available for mobile support in the future.
- In addition to Deep Research, Google introduced new models in Gemini Advanced, such as 2.0 Flash Thinking Experimental and 2.0 Pro Experimental, ranked highly in generative AI models.
- The 2.0 Flash Thinking Experimental model is Google's most advanced reasoning model, breaking down prompts into steps to enhance reasoning and understanding for users.
- The expansion of Deep Research and new AI models aim to offer powerful research and generative capabilities for Google Workspace users, with support for more languages planned.
Read Full Article
Like
Arstechnica
1M
371
Image Credit: Arstechnica
Microsoft’s new AI agent can control software and robots
- Microsoft Research has introduced Magma, an integrated AI model for controlling software interfaces and robots.
- Magma is the first AI model that can process and act upon multimodal data.
- The project is a collaboration between Microsoft and several universities.
- Unlike previous systems, Magma integrates perception and control into a single foundation model.
Read Full Article
22 Likes
The Verge
1M
137
Image Credit: The Verge
Fitbit’s got a battery problem
- Some users have reported a significant drop in battery life after a mandatory firmware update for Fitbit Versa 3 and Sense.
- The firmware update was aimed at reducing the risk of battery overheating.
- While the update fixed the overheating issue, it resulted in shorter battery life, making devices last only a day instead of 2-6 days.
- This isn't the first time Fitbit has faced battery-related complaints, and users have expressed dissatisfaction with the customer support.
Read Full Article
8 Likes
Pymnts
1M
435
Image Credit: Pymnts
Google’s AI ‘Co-Scientist’ Helps Unearth Research Ideas
- Google has introduced an artificial intelligence (AI) system called 'AI co-scientist' to help scientists come up with new hypotheses grounded in research literature.
- The AI system involves AI agents generating ideas, debating them in a 'tournament,' refining hypotheses, and presenting them to human scientists.
- AI co-scientist aims to accelerate efforts in science and medicine by producing new ideas, rather than summarizing existing ones.
- While initially applied to biomedicine, the AI system can be utilized in various fields powered by large language models like Gemini 2.0.
- The AI co-scientist complements human scientists but does not replace them, functioning as a collaborative tool requiring human input and refinement.
- The increasing use of AI in scientific research poses concerns about the authenticity and integrity of the scientific record.
- The AI co-scientist comprises multiple AI agents that interact to fulfill research goals, with actions including idea generation, clustering, and refining hypotheses.
- Researchers tested the AI co-scientist in tasks related to antibiotic resistance, demonstrating its ability to generate accurate research proposals within a short timeframe.
- The AI co-scientist's success in formulating hypotheses demonstrates its potential to enhance biomedical and scientific discovery processes.
- Google’s AI 'co-scientist' contributes to advancing research ideas and could potentially revolutionize scientific discovery in various fields.
Read Full Article
26 Likes
Discover more
Androidauthority
1M
165
Image Credit: Androidauthority
YouTube Premium Lite is coming to the US for the first time, but there’s a catch for music lovers
- YouTube will launch a Premium Lite plan for US, Australia, Germany, and Thailand.
- Premium Lite will allow ad-free video viewing, excluding music videos.
- The plan is expected to be introduced soon in the mentioned countries.
- It may eventually expand to other markets as well.
Read Full Article
9 Likes
Androidauthority
1M
119
Image Credit: Androidauthority
Your Google Assistant-integrated headphones are losing this convenient hands-free feature
- Google has disabled automatic Assistant notification reading on all headphones.
- This feature was previously disabled for Pixel Buds Pro last September.
- Users can still manually ask Assistant to read notifications.
- The move is seen as Google dialing back on Assistant features for headphones.
Read Full Article
7 Likes
Androidauthority
1M
385
Image Credit: Androidauthority
Google Home app is getting a YouTube-like feature for your video history
- The Google Home mobile app introduces a new way to navigate through video history.
- Users can double-tap to rewind or fast-forward by 10-second increments.
- This feature is available in the Public Preview program.
- Users can navigate camera history faster using the quick seek feature.
Read Full Article
23 Likes
Googleblog
1M
55
Image Credit: Googleblog
Meet the Android Studio Team: A Conversation with Engineering Director, Tor Norbye
- The 'Meet the Android Studio Team' blog series introduces Tor Norbye, an Engineering Director at Google, leading the Android Studio development.
- Tor shares his journey from Norway to Android development, starting with coding in his early days and transitioning to mobile computing at Google.
- Android Studio evolved from Eclipse-based tooling to being built on IntelliJ, enhancing the development experience for Android developers.
- The integration of AI in Android development has transformed developer capabilities, enabling the introduction of new features like intelligent code completion.
- Tor emphasizes a work-life balance for himself and the team to ensure sustainability and long-term success in their roles.
- User feedback through channels like the Android Studio issue tracker influences feature development and bug fixes, improving the overall user experience.
- The Studio team aligns with Google's vision for the Android platform by supporting the latest technologies and recommendations, such as Kotlin, Coroutines, and Material design.
- AI-driven tools in Android Studio, like automated refactoring and improved code completion, aim to enhance developers' productivity and efficiency.
- Future advancements in AI and machine learning within Android Studio will continue to empower developers with new capabilities and features.
- Stay tuned for the next installment in the series to learn more about the talented individuals shaping Android Studio's development.
Read Full Article
3 Likes
Androidauthority
1M
320
Image Credit: Androidauthority
Google’s phone app can now filter out spam and unknown calls from your call log
- Google’s Phone app is adding new filters to help users organize their call logs and reduce clutter.
- The new filters let users sort call history by missed calls, calls from contacts, and spam-marked numbers.
- Google has been expanding its Call Assist features with tools like Hold for Me, Direct My Call, and reverse phone lookup.
- The phone app feature is now officially rolling out after previously being spotted in a beta version.
Read Full Article
19 Likes
Medium
1M
243
Image Credit: Medium
Inference PaliGemma 2 with Transformers.js
- PaliGemma 2, the latest VLM in the Gemma family, improves adaptability on various transfer tasks using the SigLIP vision encoder and Gemma 2 language decoder architecture.
- The article details how to serve PaliGemma 2 with Transformers.js in the browser for making inferences directly.
- Steps include converting PaliGemma 2 model components to ONNX format, quantizing the models, and uploading the ONNX weights on Hugging Face for inference.
- Part I involves converting and quantizing the vision encoder, language decoder, and embedding tokens of PaliGemma 2, ensuring compatibility for inference using Transformers.js.
- In Part II, the process of running inferences with the converted model is demonstrated, either on Colab or in a web app using Node.js.
- A Node.js web app that accepts image and text prompts, leveraging the converted PaliGemma 2 model, is showcased in the article.
- The author shares insights on leveraging ONNX to run PaliGemma 2 in a non-Pythonic environment, highlighting the efforts and learnings involved in the project.
- Acknowledgments are given to the ONNX community and the Google ML Developers Program team for providing support and resources for the project.
- The article concludes with reference links and encourages readers to explore and engage with the contents shared.
- The demonstration includes results of inferencing with PaliGemma 2 Mix models using Transformers.js, showcasing impressive performance.
- The author encourages appreciation by starring the repository, sharing the work, and hints at future engaging blogs on related topics.
Read Full Article
14 Likes
Androidauthority
1M
13
Image Credit: Androidauthority
Google Meet has a new look on phones, and these familiar buttons are moving
- Google is rolling out design tweaks to the Meet app on Android and iOS, changing the placement of some key controls.
- Emoji reactions are moving to the bottom bar for easier access, while the hand raise button is shifting into the overflow menu on mobile devices.
- Host controls are relocating to the settings menu, and iOS users will see the end call button move to the right side.
- The design refresh aims to make it easier for users to navigate meetings on small screens without affecting the functionality of the features.
Read Full Article
Like
Androidauthority
1M
243
Image Credit: Androidauthority
YouTube Music just gave its AI playlist maker a revamp
- The YouTube Music app has undergone an update that changes how 'Ask for music' appears.
- The 'Ask for music any way you like' card has been transformed into an 'Ask for music' carousel.
- The carousel offers 10 different example prompts for generating custom AI-generated radio/playlist.
- Some of the example prompts include deep-voice singers, new releases, chillwave, British rock classics, etc.
Read Full Article
14 Likes
The Verge
1M
224
Image Credit: The Verge
Amazon’s Appstore is pulling Android support
- Amazon is ending Android support for its Appstore this year on August 20th.
- Apps downloaded from Amazon’s Appstore onto Android devices won’t guarantee operation after August 20th.
- Amazon is also discontinuing its Amazon Coins program, a virtual currency used for purchasing apps and in-app items.
- Amazon's Appstore has been unable to compete with Google's Android market dominance.
Read Full Article
13 Likes
Droid-Life
1M
32
Image Credit: Droid-Life
Gemini Advanced Now Gets ‘Deep Research’ on Mobile
- Gemini Advanced users on mobile now have access to Deep Research.
- Deep Research generates comprehensive reports on various research topics.
- It aims to save users hours of research time.
- Gemini Advanced is required to utilize Deep Research.
Read Full Article
1 Like
Cloudblog
1M
302
Image Credit: Cloudblog
An SRE’s guide to optimizing ML systems with MLOps pipelines
- Site Reliability Engineers (SREs) responsible for ML systems need to focus on pipeline management for reliable services.
- Training ML models involves managing data ingestion, freshness, training, and deployment efficiency.
- ML systems' challenges include capacity planning, resource management, and cost understanding.
- Data freshness is crucial for ML systems' health and user experience, with varying requirements by product.
- Automation, SLOs, and understanding data volume are key for reliable ML pipelines.
- Efficient machine learning inference is vital for real-world deployment, optimizing cost effectiveness.
- Knowing how to efficiently use specialized hardware is essential for cost-effective AI pipelines.
- Automation and minimizing manual effort are crucial for building resilient data pipelines.
- Successful ML deployments require holistic management, including data pipelines, training, and monitoring.
- Using GKE for AI orchestration can further enhance ML system management and optimization.
Read Full Article
18 Likes
For uninterrupted reading, download the app