– Meta releases new Llama series AI models for text analysis and writing
– Debate over the definition of open source with regards to AI models like Llama 3 by Meta
– Other recent notable AI stories include Meta chatbot update, Snap adding watermarks to AI-generated images, and Boston Dynamics unveiling all-electric humanoid Atlas robot.
This week, Meta released two new generative AI models, Llama 3 8B and Llama 3 70B, capable of analyzing and writing text. While Meta deemed these models open source, there are certain restrictions imposed that challenge this classification. The debate over what constitutes open source in the AI industry has been ongoing, particularly as big tech companies like Meta release models that come with significant limitations.
In other AI news, Meta updated its chatbot powered by Llama 3 across its platforms, Snap plans to add watermarks to AI-generated images, Boston Dynamics unveiled its new humanoid robot Atlas, and a new startup, Menteebot, focused on bipedal robotics systems. Reddit is working on an AI-powered language translation feature, while LinkedIn quietly tests AI-generated content for premium company pages.
On the research front, Swiss researchers found AI chatbots can be more persuasive in debates if equipped with personal information, raising concerns similar to Cambridge Analytica. Stanford’s Christopher Manning, an expert in language research, was awarded the John von Neuman Medal. Stuart Russell and Michael Cohen discuss the challenge of regulating advanced AIs capable of long-term planning, suggesting restrictions on hardware. Researchers also explore new neuromorphic computers, such as Sandia National Labs’ Hala Point system with 1.15 billion artificial neurons, aiming to innovate more brain-like approaches to existing algorithms. The fast-paced developments in the AI industry underscore the challenges and ethical considerations surrounding the advancement of artificial intelligence technologies.