OpenAI sets up safety committee as it starts training latest AI mode

OpenAI has announced the creation of a Safety and Security Committee and the training of a new AI model to succeed GPT-4. The committee’s role is to provide recommendations on safety and security for OpenAI’s projects. Initially, it will spend 90 days assessing and enhancing OpenAI’s safety processes before presenting its findings to the board. The board will then review and publicly share the adopted recommendations.

The establishment of this committee comes amid internal debates on AI safety, highlighted by the resignations of researcher Jan Leike and co-founder Ilya Sutskever, which led to the disbandment of the “superalignment” team focused on AI risks. OpenAI’s advanced AI chatbots, capable of human-like conversations and generating images from text, have sparked safety concerns. Additionally, OpenAI has started training a new AI model aimed at advancing capabilities towards Artificial General Intelligence (AGI).

Source – CGTN