OpenAI has disbanded its team dedicated to long-term AI risk assessment, merely a year after its inception, according to a source familiar with the matter who spoke to CNBC on Friday.
The source, speaking on condition of anonymity, revealed that some team members have been reassigned to various other departments within the company.
This development follows the recent departures of both team leaders, OpenAI co-founder Ilya Sutskever and Jan Leike, from the Microsoft-backed startup. Leike expressed concerns about OpenAI’s shift in focus from safety protocols to product development.
Initially introduced last year, OpenAI’s Superalignment team aimed to spearhead scientific and technical innovations to govern AI systems surpassing human intelligence. The initiative promised to allocate 20% of the company’s computing resources over four years.
OpenAI declined to comment directly, referring CNBC to a post by co-founder and CEO Sam Altman on X, expressing regret over Leike’s departure and acknowledging the company’s ongoing responsibilities. Additionally, OpenAI co-founder Greg Brockman posted a statement emphasizing the company’s efforts in raising awareness regarding the risks and potentials of AGI.
While Sutskever and Leike announced their exits on the social media platform X, Leike elaborated further on his reasons for leaving, citing discrepancies in priorities between himself and OpenAI leadership. He stressed the necessity for the company to prioritize security, monitoring, preparedness, safety, and societal impact.
Leike underscored the critical importance of transforming OpenAI into a “safety-first AGI company,” recognizing the inherent risks in developing machines surpassing human intelligence.
The dissolution of the Superalignment team and the high-profile departures come amid a tumultuous period for OpenAI, including a leadership crisis involving Altman in November. Altman’s removal from the board and subsequent reinstatement prompted resignations and sparked upheaval among stakeholders, including Microsoft.
Altman expressed sadness over Sutskever’s departure, praising his intellect and vision. He announced that research director Jakub Pachocki would assume Sutskever’s role as chief scientist.
These developments coincide with OpenAI’s recent launch of a new AI model and desktop version of ChatGPT, signaling the company’s ongoing efforts to broaden the application of its popular chatbot. The update introduces the GPT-4 model to all users, boasting enhanced capabilities in text, video, and audio processing.
OpenAI aims to further enhance user experience by enabling video chat functionality with ChatGPT, marking a significant advancement in usability, as highlighted by technology chief Mira Murati in a livestreamed event.
Source: cnbc.com
