🗞 Day 6 of "12 Days of OpenAI", Microsoft Launches Phi-4, and Claude 3.5

🔎 The Latest on the AI Frontier:

• Day 6 of "12 Days of OpenAI" event: ChatGPT with Screen-Sharing and Voice Features

• Microsoft Debuts Phi-4: New 14B Parameter Model Specialized in Mathematics

• Claude 3.5 Haiku Goes Public with Industry-Leading Context Window

• Google Gemini Prepares to Launch Automated Prompt Scheduling

• Samsung and Google Join Forces on 'Project Moohan' XR Headset

• Project Astra: Google's New Multimodal AI Assistant Combines Vision and Memory

• Other news you might find interesting

🎄 OpenAI launches multi-feature update on Day 6 of "12 Days of OpenAI" event

• Advanced Voice Mode gains screen-sharing and visual capabilities, allowing ChatGPT to provide real-time guidance through video and screen interactions for Team, Pro, and Plus users.

• A festive Santa voice option is rolling out across all platforms with ChatGPT voice mode, with users receiving a one-time reset of usage limits to try the feature.

• The rollout schedule varies by region and subscription tier, with European Pro and Plus subscribers gaining access later, and Enterprise/Edu users scheduled for early 2025.

🔬 Microsoft releases Phi-4, a compact 14B parameter AI model focused on mathematics performance

• The new model is exclusively available for research via Azure AI Foundry platform, competing with other small models like GPT-4o mini, Gemini 2.0 Flash, and Claude 3.5 Haiku.

• Performance improvements are attributed to high-quality synthetic datasets combined with human-generated content, addressing the industry's "pre-training data wall."

• This marks the first Phi-series launch since the departure of VP Sébastien Bubeck to OpenAI in October.

🚀 Anthropic's Claude 3.5 Haiku, their fastest AI model, is now available to all users via the Claude chatbot on web and mobile platforms

• With a 200,000-token context window, it surpasses OpenAI's GPT-4o's 128,000 tokens, enabling efficient processing of extensive inputs.

• Optimized for real-time tasks, Claude 3.5 Haiku excels in handling large datasets, analyzing financial documents, and generating outputs from long-context information.

• Users can access it for free with a variable daily message limit or subscribe to the Claude Pro plan at $20 per month for increased usage and additional features.

⚡ Google's Gemini to introduce scheduled prompts feature for automated AI interactions

• The development includes a functional UI allowing users to schedule, modify timing, and manage frequency of automated prompts through a dedicated settings menu.

• The feature is currently hidden behind a feature flag but appears to be fully operational, suggesting an imminent release.

• This rollout aligns with Google's December product launch strategy, which has seen multiple new features and products being released this month.

🕶️ Samsung and Google unveil 'Project Moohan,' a mixed-reality headset powered by the new Android XR operating system, aiming to challenge Apple's Vision Pro

• Set to launch in 2025, 'Project Moohan' will feature both virtual and augmented reality capabilities, integrating Google's Gemini AI for natural language interactions.

• The headset is expected to utilize Qualcomm's Snapdragon XR2+ Gen 2 processor, supporting high-resolution displays and multiple camera inputs for advanced tracking.

• Samsung's collaboration with Google and Qualcomm reflects a strategic move to establish a foothold in the competitive XR market, leveraging AI advancements to enhance user experience.

🌟 Google launches Project Astra, showcasing advanced multimodal AI assistant that combines vision, voice, and memory

• The system can interpret real-time video, respond to voice commands, and remember context from previous interactions, powered by the newly announced Gemini 2.0 model.

• During a live demo, Astra demonstrated practical applications like reading recipes, identifying objects, and making recommendations, though still showing occasional glitches that required verbal corrections.

• Despite impressive capabilities, Google hasn't announced a release date, and experts raise concerns about privacy implications and the need for more transparency about how the system works.

🖨 More news you might find interesting:

• OpenAI's Sam Altman reportedly pledges $1M to Trump's inaugural fund alongside other tech giants.

• Apple introduces 'Genmoji,' an AI-powered feature that allows users to create personalized emojis using text descriptions or photos.

• Texas Attorney General launches investigation into 14 online platforms over child safety concerns.

• Google announces Android XR operating system for next-generation mixed reality headsets.

• Researchers uncover new AI model theft technique using electromagnetic signals with 99.91% accuracy.

• NYU study finds AI models exhibit human-like "us vs. them" social biases.

• Scale AI faces class action lawsuit over alleged worker exploitation and misclassification.

• Amazon launches vehicle sales on its platform through exclusive partnership with Hyundai.

• ACLU warns against Axon's AI police report system, citing major civil liberties concerns.

• Russia announces BRICS-centered AI Alliance Network to challenge Western dominance.

• SEC plans new charges against Elon Musk and investigates Neuralink, lawyer reveals.

Have any feedback? Send us an email