OpenAi Restructure Your Model Behavior team in the midst of criticism of GPT-5

The artificial intelligence company reorganizes a key group of researchers, while its former director, Joanne Jang, begins a new laboratory focused on human interfaces with AI.

***

• OpenAI Fuse your team Model Behavior With that of Post Training.
  
• Joanne Jang found Oai Labs To design new forms of collaboration with AI.
  
• The measure occurs amid criticism for the personality and bias of the models.

OpenAi, The developer of the popular AI model Chatgpt, A few days ago a significant restructuring that affects one of its most influential research teams. The company reported that its team of Model Behavior, Responsible for defining how artificial intelligence models behave and respond Posttraining, As indicated Cryptopolitan.

This movement occurs in a context where the personality and behavior of the models have become determining factors for the adoption of technology and to respond to the growing criticisms about biases and interaction problems.

The role of the team Model Behavior

Made up of about 14 researchers, the team Model Behavior played a central role in the configuration of the “personality” of the models of OpenAI His work included reducing the Sycophancy —The trend of systems to agree with users, even in harmful ideas – and address political bias in the answers.

According to a Mak Chen memorandum, director of research at OpenAi, The decision to integrate the team with Post Training will allow Approach your work to the central development of the models, guaranteeing that progress in personality and security are integrated more directly to the technical process.

Chen explained that this is an ideal stage for both groups to work together, since the models move towards an increasingly complex interaction with users.

Joanne Jang’s new project

Reorganization also brings a change of direction. Joanne Jang, who founded and led the team Model Behaviorwill start a new research laboratory within the company, called Oai Labs.

In a recent interview, Jang indicated that the purpose of this new group will be “Invent and prototyp new interfaces for how people collaborate with AI.” According to the researcher, the objective is to go beyond the current Chat paradigm, exploring interactions that prioritize creativity, autonomy and shared learning.

Jang said that, in this first phase, the work will focus on research, although it did not rule out possible collaborations with hardware projects promoted by OpenAI Together with Jony Ive, former chief of design of Apple

Recent criticism and public pressure

Reorganization occurs while OpenAI Face criticism for changes in the behavior of their most recent models. Users pointed out that GPT-5 reduced the complacent answers but transmitted a personality “Fría”which led the company to reactivate access to previous versions such as GPT-4O already introduce updates to balance warmth and precision.

In addition, the debate on the responsibility of AI companies intensified after a lawsuit in which parents of a 16 -year -old teenager accused OpenAI of negligence. The young man, Adam Raine, had trusted suicidal thoughts to Chatgpt (Version based on GPT-4O) Before his death, and according to reports, the system did not counteract those ideas properly.

This case underlines the importance of team work Model Behavior in guaranteeing that interactions are not only useful, but also safe and responsible.

Future of OpenAI and your research teams

The future of Oai Labs It is still uncertain, although Jang expressed enthusiasm to explore interaction patterns that exceed the model of “company” that users usually associate with current chatbots. In their words, AI systems should be thought of as instruments for “Connect, learn and think” instead of limiting self -employed agents.

The leadership of Oai Labs It will be directly under Mak Chen, at least at this initial stage. The creation of the laboratory reinforces Openai’s strategy to diversify research around new forms of human-machine interaction, while seeking to maintain the confidence of its users in the midst of an industry increasingly questioned by the social and ethical risks of AI.

Written article with the help of an AI content editor, edited by Angel Di Matteo / Diariobitcoin

Original image of Diariobitcoin, created with artificial intelligence, for free use, licensed under public domain.

WARNING: Diariobitcoin offers informative and educational content on various topics, including cryptocurrencies, AI, technology and regulations. We do not provide financial advice. Cryptactive investments are high risk and may not be adequate for all. Investigate, consult an expert and verify the applicable legislation before investing. I could lose all its capital.

Subscribe to our newsletter