OpenAI Shakes Up Team Behind ChatGPT's Personality in Major Reorganization
In a significant move, OpenAI has reorganized its Model Behavior research team, which focuses on developing AI models' personalities and interactions with humans, by merging it with the Post Training team. According to an August memo seen by TechCrunch, the 14-strong Model Behavior team will now report to Max Schwarzer, the lead of the Post Training team.
The changes aim to integrate the team's expertise in shaping AI personalities and reducing sycophancy into broader efforts to improve OpenAI models' capabilities. As part of the reorganization, founding leader Joanne Jang is moving on to start a new project called OAI Labs, which will explore novel interfaces for human-AI collaboration.
According to an interview with TechCrunch, Mark Chen, OpenAI's chief research officer, stated that the Model Behavior team would be joining the Post Training team "to leverage their expertise in shaping AI personalities and interactions with humans." Chen added that the move is part of a larger effort to improve OpenAI models' capabilities.
The reorganization comes as OpenAI continues to push the boundaries of AI development. The company's flagship model, ChatGPT, has gained widespread attention for its ability to engage in natural-sounding conversations. However, concerns have been raised about the potential risks associated with creating AI models that can mimic human personalities and interactions.
According to an OpenAI spokesperson, the changes are intended to "foster a more collaborative and integrated research environment." The spokesperson added that the reorganization will allow for greater sharing of expertise and resources across teams.
The Model Behavior team's move to the Post Training team has been met with mixed reactions from industry experts. Some have praised the decision as a necessary step towards improving AI models' capabilities, while others have expressed concerns about the potential loss of specialized knowledge and expertise.
As OpenAI continues to navigate the complex landscape of AI development, it remains to be seen how the reorganization will impact the company's research efforts. However, one thing is clear: the move marks a significant shift in OpenAI's approach to developing AI models that can interact with humans in a more natural and effective way.
In related news, Joanne Jang, the founding leader of the Model Behavior team, will be leading OAI Labs, a new project exploring novel interfaces for human-AI collaboration. According to an OpenAI spokesperson, the project aims to "push the boundaries of what is possible in human-AI interaction."
This story was compiled from reports by TechCrunch and TechCrunch.
