Formation of New Safety Committee
OpenAI has created a new safety committee to oversee the training of its latest artificial intelligence (AI) model. Led by CEO Sam Altman, the committee will provide recommendations on critical security decisions to the company’s board. The formation of this committee underscores OpenAI’s commitment to enhancing safety measures as it develops more advanced AI technologies.
Leadership and Initial Tasks
The committee is chaired by Sam Altman, along with Bret Taylor, Adam D’Angelo, and Nicole Seligman. Their first task is to review and update OpenAI’s current safety practices within the next 90 days, after which they will present their recommendations to the board. Once approved, these recommendations will be made public to ensure transparency and accountability.
Context of the Committee’s Formation
This new safety initiative follows the recent disbanding of OpenAI’s Superalignment Team, which was initially tasked with addressing AI risks but was dissolved after less than a year due to the departure of its co-chairs. The establishment of the safety committee marks a renewed effort to address potential AI threats and ensure robust safety protocols.
Recent Safety Measures and AI Model Developments
OpenAI recently issued a “safety update” post-AI Seoul Summit, pledging not to release any new AI models that pose more than a “medium” threat level, based on internal evaluations. The company also plans to introduce additional protections for children and a new tool to identify AI-generated images by DALL-E 3. These measures reflect OpenAI’s proactive stance on mitigating potential risks associated with AI advancements.
Expanding the Committee with Expert Advice
The safety committee includes tech and policy experts Aleksander Madry, Lilian Weng, and newly appointed chief scientist Jakub Pachocki. Additionally, former cybersecurity officials Rob Joyce and John Carlin will provide advisory support. OpenAI’s latest model, GPT-4 Omni, aims to enhance human-computer interaction by processing audio, vision, and text inputs in real-time, signifying a significant leap in AI capabilities while maintaining a focus on safety.
SOURCE: Ref Image from Reuters
Whether writing about complex technical topics or breaking news stories, my writing is always clear, concise, and engaging. My dedication to my craft and passion for storytelling have earned me a reputation as a highly respected article writer.