Meta plans to utilize public content and user interactions from EU users to train its AI models for better local language and cultural understanding, while offering an opt-out option to comply with regulations and enhance transparency.
Researchers developed a multimodal AI tool that combines language models and graph-based models to efficiently design new molecules with desired properties from natural language queries, outperforming existing methods and potentially automating the entire molecule design and synthesis process.
BrowseComp, a novel AI benchmark, evaluates web browsing skills by posing 1,266 complex queries requiring synthesis of information across multiple websites, with a deep research model achieving 51.5% accuracy, surpassing other models' performance on this challenging task.
Meta's open-source Maverick AI trails rivals like GPT-4 and Claude on chat benchmarks, but the company hopes developers will customize the model for diverse applications, leveraging its conversational capabilities.
MonkeysPaw, a Ruby web framework, conjures entire web pages from natural language prompts, handling HTML, CSS, JavaScript, routing, and layouts, enabling developers to focus on describing desired outcomes rather than coding intricacies, albeit with potential trade-offs in performance and precision.
Elon Musk's AI venture xAI enters the market with the launch of Grok 3 API, offering a large language model and a smaller version at competitive pricing, though with a reduced context window.
Anthropic's new Max plan unlocks higher usage limits, priority access to new features and models, and enhanced capabilities for frequent Claude users who require extended conversations, complex data handling, and heavy reliance on the AI assistant throughout their day.
Google aims to merge its AI models Gemini and Veo, combining Gemini's multimodal capabilities with Veo's understanding of real-world physics learned from YouTube data, to create an enhanced universal digital assistant with improved physical world comprehension.
Multiverse Computing's CompactifAI technology shrinks Llama AI models by 80%, boosting efficiency with 60% fewer parameters, 84% greater energy efficiency, 40% faster inference, and 50% cost reduction, enabling new use cases for portable AI.
CompactifAI unveils LLaMA 3.3 70B Slim, a compressed AI model 80% smaller, 2.18x faster, with near-perfect accuracy, enabling edge deployment, lower energy consumption, and secure enterprise AI solutions while maintaining industry-leading performance.