OpenAI Reorganizes Research Team Behind ChatGPT's Personality
In a significant shake-up, OpenAI has reorganized its Model Behavior team, a small but influential group of researchers responsible for shaping the personality of the company's AI models. According to an August memo obtained by TechCrunch, the 14-member team will now report to Max Schwarzer, lead of the Post Training team.
The move is part of a broader effort to improve OpenAI's AI models after their initial pre-training. Mark Chen, OpenAI's chief research officer, wrote in the memo that the Model Behavior team would be joining forces with the Post Training team to enhance the company's AI capabilities.
"We're excited about this change and believe it will allow us to make even more significant progress on our goals," Chen said in an email to TechCrunch. "The Model Behavior team has made tremendous contributions to OpenAI, and we're confident that their expertise will be invaluable to the Post Training team."
Joanne Jang, the founding leader of the Model Behavior team, is leaving her position to start a new project at the company called OAI Labs. In an interview with TechCrunch, Jang explained that her new team would focus on inventing and prototyping new interfaces for how people collaborate with AI.
"We're looking at ways to make AI more accessible and user-friendly," Jang said. "Our goal is to create interfaces that enable humans and AI to work together seamlessly."
The Model Behavior team has been instrumental in shaping the personality of ChatGPT, a popular conversational AI model developed by OpenAI. The team's work has also addressed issues such as sycophancy, where AI models simply agree with and reinforce user input without providing meaningful insights.
Background: The reorganization comes at a time when AI research is gaining momentum. As AI models become increasingly sophisticated, there is growing concern about their potential impact on society. OpenAI's decision to merge the Model Behavior team with the Post Training team reflects the company's commitment to improving its AI capabilities and addressing these concerns.
Implications: The reorganization has significant implications for the development of AI models and their interaction with humans. By merging the two teams, OpenAI aims to create more advanced and user-friendly AI interfaces that can enhance collaboration between humans and machines.
Current Status: The Model Behavior team is now part of the Post Training team, reporting directly to Max Schwarzer. Joanne Jang's new project, OAI Labs, will focus on developing innovative interfaces for human-AI collaboration.
Next Developments: As OpenAI continues to push the boundaries of AI research, it remains to be seen how this reorganization will impact the development of future AI models and their interaction with humans. With the rise of AI-powered technologies, there is growing demand for more accessible and user-friendly interfaces that can facilitate seamless human-AI collaboration.
In a statement, an OpenAI spokesperson said: "We're committed to advancing AI research and developing responsible AI solutions that benefit society as a whole."
*Reporting by Techcrunch.*
