News/Leaks & Rumors
Gemini may get an ‘I’m feeling lucky’ button
Artificial intelligence leader Google is preparing to refresh its Gemini AI interface with new features that harken back to classic Google Search functionality, including randomized prompt suggestions and an "I'm feeling lucky" button. Key updates: Google's latest beta version of the Gemini app reveals plans to reintroduce suggestion features while maintaining a streamlined user experience. The company is bringing back randomized prompt suggestions that were previously removed from the Gemini homepage in October A new "I'm feeling lucky" button, similar to the iconic Google Search feature, will be permanently positioned first among the suggestions When tested, the "I'm feeling lucky"...
read Nov 28, 2024Google is training Gemini to manage large-scale code processing
The evolution of AI coding assistants continues as Google prepares to enhance Gemini's code analysis capabilities, potentially transforming how developers interact with and understand complex codebases. Key development: Google is preparing to upgrade Gemini to analyze entire folders of code simultaneously, moving beyond its current single-file limitation. The upcoming feature will allow users to upload up to 1,000 files totaling 100MB in a single folder This capability matches existing features offered by competitors like ChatGPT Developers will be able to query Gemini about the code's functionality and potential improvements Technical implications: The folder analysis capability represents a significant enhancement to...
read Nov 27, 2024OpenAI shuts down access to Sora after leak by protesting artists
The relationship between AI companies and artists continues to be strained as controversy erupts over OpenAI's early access program for its text-to-video generation tool, Sora. Initial controversy: OpenAI's early access program for Sora, granted to approximately 300 visual artists and filmmakers, sparked significant backlash when testers publicly released the tool along with a protest manifesto. A group of 19 artists posted their criticism on AI development site Hugging Face, leading OpenAI to suspend access within three hours The artists accused OpenAI of exploiting them for "art washing" and free labor, despite the company's $157 billion valuation The protest group emphasized...
read Nov 27, 2024What we know about the artists who leaked OpenAI’s Sora model
The tension between AI companies and artists has reached a new flashpoint as OpenAI faces backlash from testers of its Sora video generation platform. The incident unfolds: A group of artists calling themselves PR Puppets briefly shared public access to OpenAI's Sora video generation platform, leading to an immediate suspension of the testing program. The group posted a "Generate with Sora" access point on Hugging Face, allowing public access to OpenAI's actual Sora API Social media users quickly confirmed the authenticity of the access through the videos.openai.com domain OpenAI revoked access within hours, though many users managed to generate and...
read Nov 27, 2024The latest on the Sora leak saga
The race to develop and safely deploy AI video generation technology has hit a speed bump with a significant leak of OpenAI's unreleased Sora platform. The security breach: A group of artists with early access to OpenAI's Sora AI video generator leaked their access keys on HuggingFace, prompting an immediate shutdown of the preview program. The leak occurred when artists who felt they were being used as "PR puppets" shared their access credentials in protest This unauthorized sharing forced OpenAI to revoke all early access permissions to protect the platform's integrity The incident highlights the delicate balance between gathering user...
read Nov 26, 2024Artist beta testers leak OpenAI’s Sora video model in protest
The emergence of powerful AI text-to-video technology has sparked controversy as artists protest OpenAI's development practices for its Sora model. Breaking developments: A group of Sora beta testers claims to have leaked early access to OpenAI's text-to-video AI model in protest of what they describe as exploitative testing practices. The alleged leak was posted on Hugging Face, a public AI model repository, allowing users to briefly generate AI videos similar to OpenAI's official demos before access was terminated Beta testers published an open letter criticizing OpenAI's approach to artist collaboration and testing OpenAI has not confirmed whether the leak was...
read Nov 24, 2024The latest updates on Apple’s rumored ‘LLM Siri’
The rapid evolution of AI assistants has put pressure on Apple to modernize Siri, leading to reports of a significant overhaul planned for 2026. Key Development: Apple is developing a new version of Siri powered by large language models (LLMs), technology that enables AI systems to process and respond to human language more naturally. Internal testing is already underway on iPhones, iPads, and Macs through a standalone app dubbed "LLM Siri" The new AI assistant could be announced in 2025 as part of iOS 19 and macOS 16, with a planned launch in spring 2026 This timeline mirrors Apple's gradual...
read Nov 22, 2024ChatGPT may soon get a ‘Live Camera’ feature — here’s what we know
The imminent addition of real-time video capabilities to ChatGPT marks a significant expansion of the AI chatbot's sensory capabilities, moving beyond text to include visual understanding and analysis. Latest developments: OpenAI's ChatGPT mobile app beta code reveals plans for an upcoming 'Live Camera' feature that will integrate with its Advanced Voice Mode. Code snippets in version 1.2024.317 indicate functionality for real-time video processing and visual recognition The feature will enable users to engage with ChatGPT while using their device's camera for live visual feedback Integration with Advanced Voice Mode suggests a multi-modal interaction combining visual and audio inputs Technical capabilities:...
read Nov 22, 2024Android 16 may give Gemini agentic powers over mobile apps
The upcoming Android 16 operating system could significantly expand the capabilities of Google's AI assistant Gemini, potentially enabling it to control smartphone apps in ways that were previously impossible. Core functionality: Android 16 introduces a new system-level API that will allow authorized apps to perform actions within other applications on behalf of users. The feature, called "app functions," creates a framework for apps to expose specific functionalities to the Android system This capability will be restricted to system apps with special permissions, with Gemini likely to become the default assistant on new Android devices The new API integrates with Android's...
read Nov 22, 2024OpenAI is considering developing a custom web browser
Breaking OpenAI News: OpenAI is exploring the possibility of developing its own web browser, marking a potential expansion beyond its current AI offerings like ChatGPT and search capabilities. Strategic expansion details: The development of a proprietary web browser would represent a significant step forward in OpenAI's growing ecosystem of AI tools and services. The initiative appears to be in early exploration phases, with no confirmed timeline for development or release This move would complement OpenAI's existing portfolio, which includes ChatGPT and its AI-powered search functionality A browser could potentially provide a more integrated and seamless experience for users of OpenAI's...
read Nov 18, 2024The Google smart home ecosystem may get a big Gemini AI upgrade
Google's expanding influence in artificial intelligence technology is set to make its smart home assistant more conversational and intuitive through integration with its Gemini AI model. Key Developments: Recent analysis of Google's app code reveals significant planned upgrades to Google Assistant's functionality and user interaction capabilities. Code snippets indicate the introduction of an improved Assistant with enhanced response capabilities and better contextual understanding The upgrade aims to make interactions feel more natural and less robotic, allowing for more fluid conversations New voice options will be available to create a more authentic speaking experience through smart speakers and displays Technical Improvements:...
read Nov 17, 2024When to expect Gemini integration with Apple Intelligence
Artificial intelligence integration for Apple's mobile operating system continues to evolve with new partnerships and features on the horizon. Current state of AI integration: Apple's partnership with OpenAI has already brought ChatGPT functionality to iOS 18.2 beta, with a full release expected in December. The integration represents Apple's first major AI chatbot partnership iOS 18.2 beta users can currently test the ChatGPT features This development follows Apple's strategy of partnering with established AI companies rather than building everything in-house Google Gemini timeline: According to Bloomberg's Mark Gurman, the integration of Google's Gemini AI into Apple Intelligence will likely be delayed...
read Nov 16, 2024Google’s AI Overviews may get a little better with this new feature
Artificial Intelligence is becoming increasingly integrated into Google Search, with new features being tested to enhance the AI Overview summaries first introduced earlier this year. Key development: Google is testing a feature that allows users to select text within AI Overviews to generate additional, nested AI-generated summaries focused on the selected content. The experimental feature was discovered in version 15.45.33 beta of the Google app on a Pixel 7a Users can highlight specific portions of an AI Overview to receive more detailed information about that particular topic The nested summaries appear in a resizable card that allows users to still...
read Nov 13, 2024OpenAI to launch task-performing AI agents ‘Operator’ in January
The artificial intelligence industry is poised for a significant shift as OpenAI prepares to launch "Operator," a new AI system capable of performing independent tasks on behalf of users, marking a transition from conversational AI to action-oriented automation. The breakthrough explained: OpenAI's upcoming "Operator" platform, scheduled for release in January 2025, represents a fundamental evolution in AI capability, moving beyond chat-based interactions to autonomous task execution. The system will be able to perform complex tasks independently, from writing code to making travel arrangements Unlike current AI models that require specific prompts, Operator will function more like a personal assistant, managing...
read Nov 13, 2024Xiaomi 15 global version spotted on Geekbench with new specs
The Xiaomi 15 series continues to make waves in the smartphone market, with new details emerging about both the global version of the base model and the Ultra variant's camera capabilities. Latest developments: The global version of the Xiaomi 15 has appeared on Geekbench's AI test listing, while new information about the Xiaomi 15 Ultra's camera system has surfaced. The global Xiaomi 15, identified by model number 24129PN74G, achieved maximum scores of 2,519 in single-precision and 2,522 in half-precision tests The device's quantized score reached 4,254, though notably these tests were conducted on CPU rather than NPU or GPU Key...
read Nov 12, 2024Apple may launch an AI-powered smart home camera by 2026
The smart home market continues to evolve as major tech companies vie for dominance in connected home devices, with Apple now planning to expand its hardware offerings beyond speakers and displays. Strategic expansion plans: Apple is reportedly preparing to enter the smart home camera market in 2026, marking its first direct competition with third-party HomeKit camera manufacturers. Ming-Chi Kuo, a prominent Apple analyst, indicates the company aims to achieve annual shipments in the tens of millions for its IP camera The device will feature deep integration with Siri and Apple Intelligence features This move represents Apple's first venture into first-party...
read Nov 12, 2024X is testing a free version of Grok that doesn’t require a blue checkmark
Latest developments: X has begun testing a free tier of its Grok AI chatbot, marking a shift from its previous subscription-only model. Select X users have received invitations to try a limited version of Grok without requiring a premium subscription The free tier includes basic chat functionality, image generation, and photo analysis capabilities Initial rollout appears to be region-specific, suggesting a gradual implementation strategy Technical specifications and limitations: The free version of Grok comes with specific usage restrictions that differentiate it from the full subscription service. Users can ask 10 Grok 2 questions and 20 Grok 2 mini questions every...
read Nov 12, 2024Samsung’s AI-powered One UI 7 beta is improving, but don’t get too excited
One UI 7, Samsung's next major software upgrade for Galaxy devices, is showing signs of development progress while maintaining an extended timeline that stretches into 2025. Initial developments and timeline: Samsung's test servers have revealed new One UI 7 beta builds, suggesting an imminent beta program launch for flagship devices. Beta firmware versions ending in ZXKB have been spotted on servers for the Galaxy S24 Ultra in Europe and India The stable version of One UI 7 is officially scheduled to debut with the Galaxy S25 in early 2025 Multiple sources indicate beta testing could begin as soon as next...
read Nov 12, 2024Slack might soon use AI to summarize your files
The workplace messaging platform Slack appears to be expanding its artificial intelligence capabilities with a new feature that would automatically generate summaries of uploaded files, building on its existing AI-powered tools introduced earlier this year. Key development: Slack's Android application code reveals plans to implement AI-powered file summarization capabilities, allowing users to quickly grasp the contents of documents without reading them in full. The feature appears designed to work with PDFs and various text document formats Certain limitations will apply, including restrictions on password-protected files and size constraints Users will have the ability to delete summaries and provide feedback on...
read Nov 11, 2024Gemini Live may start to answer question about your files soon
The evolution of Google's Gemini AI continues as new features emerge to enhance file handling and user interaction capabilities. Latest developments: Google is developing functionality that will allow users to discuss uploaded files through Gemini Live, expanding beyond the current file handling capabilities of Gemini Advanced. Current Gemini Advanced users can upload and interact with various file types, including text documents and spreadsheets The new feature will suggest transitioning to Gemini Live when users upload files The functionality aims to leverage Gemini Live's conversational nature for more dynamic file interactions Technical insights: Recent discoveries in the Google app beta version...
read Nov 11, 2024Instagram tests AI-generated profile pictures for users
Instagram is testing new AI capabilities that would allow users to generate profile pictures directly within the app, marking another step in Meta's integration of artificial intelligence across its social media platforms. Key development: Meta's Instagram platform is exploring a feature that would let users create AI-generated profile pictures without leaving the app. Reverse engineer Alessandro Paluzzi discovered a hidden menu option labeled "create An AI profile picture" within Instagram's code The feature would simplify the process for users who want to use AI-generated images as their profile pictures, eliminating the need to use external AI image generators While users...
read Nov 11, 2024New report suggests OpenAI’s Orion model faces major bottlenecks
The development of artificial intelligence models is facing unexpected hurdles as tech companies encounter diminishing returns in performance gains, highlighting broader challenges in the field of machine learning. Current challenges: OpenAI's next-generation AI model Orion is experiencing smaller-than-anticipated performance improvements compared to its predecessor GPT-4. While Orion shows enhanced language capabilities, it struggles to consistently outperform GPT-4 in specific areas, particularly in coding tasks The scarcity of high-quality training data has emerged as a significant bottleneck, with most readily available data already utilized in existing models The anticipated release date for Orion has been pushed to early 2025, and it...
read Nov 11, 2024OpenAI’s Orion model is reportedly only somewhat better than GPT-4
The development of advanced language models appears to be reaching a plateau, with OpenAI's latest model showing only modest improvements over its predecessor, highlighting broader challenges in AI advancement. Key developments: OpenAI's upcoming "Orion" model demonstrates smaller performance gains compared to the leap between GPT-3 and GPT-4, while showing improvements primarily in language capabilities. The new model may be more expensive to operate in data centers than previous versions Performance improvements in areas like programming have been inconsistent The quality gap between Orion and GPT-4 is notably smaller than expected Training data challenges: OpenAI faces limitations in accessing high-quality training...
read Nov 11, 2024Google’s Gemini AI app may soon arrive on iPhone
The potential arrival of Google's Gemini app on iPhone marks a significant expansion in the AI chatbot landscape, potentially offering iOS users a powerful alternative to existing AI assistants. Latest developments: Google is reportedly preparing to launch a dedicated Gemini app for iPhone users, with early sightings reported in the Philippines App Store. A dedicated app would provide faster access to Gemini's features compared to the current Google App integration The new app is expected to include Gemini Live, Google's conversational AI feature Currently, iPhone users can only access Gemini through the Google App, with a UI that differs from...
read