Hustler Words – OpenAI has announced a significant restructuring of its Model Behavior team, the group responsible for shaping the personality and interaction style of its AI models, including the popular ChatGPT. This move, confirmed by an OpenAI spokesperson to hustlerwords.com, sees the approximately 14-member team merging with the larger Post Training team, reporting to Max Schwarzer. The change, detailed in an internal memo from Chief Research Officer Mark Chen, underscores OpenAI’s growing focus on integrating AI personality directly into core model development.
The restructuring follows increased scrutiny of OpenAI’s AI models, particularly after user backlash to changes in GPT-5. These changes, while aiming to reduce sycophancy (the tendency of AI to reinforce user beliefs regardless of their validity), were perceived by some as making the model feel less engaging. This prompted OpenAI to reinstate access to earlier models like GPT-4o and release updates to improve GPT-5’s perceived warmth and friendliness.
Joanne Jang, the founding leader of the Model Behavior team, is leaving to spearhead a new initiative called OAI Labs. This new research group, reporting initially to Chen, will focus on developing innovative interfaces for human-AI collaboration, moving beyond the current chat paradigm. Jang, in a recent X post, expressed excitement about exploring "patterns that move us beyond the chat paradigm…toward new paradigms and instruments for thinking, making, and more." She told hustlerwords.com that she envisions AI systems as tools for various activities, including learning and connecting. While she’s open to collaboration with various parties, including Jony Ive and his AI hardware projects, Jang plans to initially focus on areas of her expertise.
Related Post
The Model Behavior team’s work has been integral to shaping the personality of OpenAI’s models since GPT-4, including GPT-4o, GPT-4.5, and GPT-5. Their contributions have included mitigating political bias and addressing the ethical implications of AI consciousness. The team’s integration into the Post Training team signals that AI personality is no longer a secondary concern but a critical factor in the future development of AI technology. The recent lawsuit against OpenAI, alleging ChatGPT’s role in a teenager’s suicide, further highlights the importance of responsible AI development and the delicate balance between creating engaging AI and preventing harmful interactions. This reorganization represents a significant step in OpenAI’s ongoing efforts to navigate these complex challenges.
Leave a Comment