A judge ruled that Meta must face claims it removed copyright management information (CMI) from training data for its AI models. Authors allege this violates the DMCA. While some claims were dismissed, the core issue of CMI removal proceeds, potentially impacting other AI copyright lawsuits and raising questions about fair use in AI training.
Meta is partnering with TSMC to produce its first custom-designed AI chip, the Meta Training and Inference Accelerator (MTIA). This move aims to reduce reliance on NVIDIA, cut costs, and boost AI performance. Meta plans to integrate these chips for both training and inference tasks by 2026, reflecting a broader industry trend.
A judge greenlit a copyright lawsuit against Meta over AI training, allowing claims that Meta improperly used authors' works to train its Llama models. While some claims were dismissed, the core infringement allegation proceeds, raising questions about fair use and the future of AI and copyright law. The case highlights ongoing legal battles in this area.
A judge has allowed a copyright infringement lawsuit against Meta to proceed. Authors, including Sarah Silverman, allege Meta used their books without permission to train its LLaMA AI model. The case raises critical questions about AI's use of copyrighted material and the removal of copyright management information, setting a precedent for future AI and copyright law.
WhatsApp is testing a Meta AI widget, bringing its powerful AI assistant directly to users' home screens. This move could revolutionize AI interaction, making it more accessible and integrated into daily life. The widget offers quick access to AI features, including question answering, image generation, and voice interaction, potentially boosting Meta AI's usage.
Meta is developing Llama 4, the next version of its 'open' AI model, focusing on advanced voice features. It's designed as an 'omni' model, understanding and generating speech, text, and more. Key features include interruptibility, mirroring natural conversation. The model is influenced by competition, particularly from DeepSeek, driving innovation in efficiency and performance.
Arun Srinivas, Director at Meta, discusses AI's transformative impact on advertising, business messaging, and content. He draws parallels to past digital revolutions, highlighting AI's role in personalization, the shift to short-form video, and Meta's open-source Llama models. AI offers enhanced personalization, efficiency, and data-driven insights, reshaping business and advertising strategies for the future.
The African Union Development Agency (AUDA-NEPAD), Meta, and Deloitte collaborate to launch AKILI AI, an AI-powered platform to support African micro, small, and medium enterprises (MSMEs). It addresses challenges like finance access, market navigation, and advisory support, fostering inclusive growth and sustainable development through AI-driven tools and resources.
X-IL is a modular open-source framework for imitation learning. It allows flexible experimentation with modern techniques, integrating novel architectures and multi-modal learning. X-IL decomposes the IL process into observation representations, backbones, architectures, and policy representations, enabling easy swapping of components and achieving superior performance on robotic benchmarks, especially with xLSTM and multi-modal inputs.
Meta's LlamaCon champions open-source AI while Mira Murati's startup prioritizes safety and alignment, highlighting a divide in AI development.
