Facebook aims to solve AI models' 'big problem' thatt made Sundar Pichai say 'We Got it Wrong'

Meta, the parent company of Facebook, has launched the Llama 4 AI model to address bias in large language models, particularly on politically and socially debated topics. The company aims to mitigate bias by making Llama 4 more responsive and balanced in its responses. Facebook claims that Llama 4 demonstrates advancements in reducing bias compared to its predecessor, Llama 3, and is now comparable to Grok, developed by Elon Musk-owned xAI.

Full Article