OpenAI Shuts Down AI Safety Team Just After Less Than a Year in Office

By

US-TECHNOLOGY-AI-ALTMAN
(ILLUSTRATION) This illustration photo produced in Arlington, Virginia on November 20, 2023, shows a smart phone screen displaying the logo of OpenAI juxtaposed with a screen showing a photo of former OpenAI CEO Sam Altman attending the Asia-Pacific Economic Cooperation (APEC) Leaders' Week in San Francisco, California, on November 16, 2023. Hundreds of staff at OpenAI threatened to quit the leading artificial intelligence company on November 20, 2023 and join Microsoft. They would follow OpenAI co-founder Sam Altman, who said he was starting an AI subsidiary at Microsoft following his shock sacking on November 17, 2023 from the company whose ChatGPT chatbot has led the rapid rise of artificial intelligence technology. In a letter, some of OpenAI's most senior staff members threatened to leave the company if the board did not get replaced. Reports said as many as 500 of OpenAI's 770 employees signed the letter. The startup's board sacked Altman on Friday, with US media citing concerns that he was underestimating the dangers of its tech and leading the company away from its stated mission -- claims his successor has denied. Photo by OLIVIER DOULIERY/AFP via Getty Images

OpenAI has shut down its team focused on the long-term risks of artificial intelligence less than a year after setting it up, a source confirmed to CNBC.

According to the anonymous source, team members face reassignment to other company teams.

The AI safety group called the 'Superalignment team,' formed last year, was dedicated to achieving breakthroughs in controlling AI systems more advanced than humans. OpenAI initially committed 20% of its computing power to this initiative over four years.

Interestingly, this decision follows the recent departures of team leaders, OpenAI co-founder Ilya Sutskever and Jan Leike, from the Microsoft-backed startup, as reported by VCPost, and new AI models and updates, including an enhanced version of ChatGPT.

The latest model, GPT-4o, offers improved performance and capabilities in text, video, and audio. Video chat features will be introduced in the future.

Concerns on OpenAI

Leike, in a recent statement also reported by CNBC, expressed concerns about OpenAI's shifting priorities, noting that the push for new products had overshadowed the company's focus on safety.

He emphasized the need for a stronger emphasis on security, monitoring, preparedness, and societal impact within OpenAI's operations.

Leike described recent months as challenging, with his team facing challenges securing the necessary research resources. He also believes that OpenAI should become a "safety-first AGI company" due to the inherent risks of developing advanced AI.

Altman recently expressed sadness over Sutskever's departure, calling him "one of the greatest minds of our generation" and a "dear friend." Jakub Pachocki, a research director at OpenAI since 2017, will succeed Sutskever as chief scientist.

Why the AI Safety Team Was Shut Down

No clear details have been shared yet on why OpenAI has made this decision. But prior to this, the company had numerous internal conflicts in November when its board of directors decided to remove CEO Sam Altman from his position. This led to employees either resigning or threatening to resign and investors, notably Microsoft, expressing their dissatisfaction.

Due to the intense reaction and pressure, Altman was quickly reinstated as CEO within a week, and several board members who supported his removal resigned.

Tags
OpenAI, AI, Artificial intelligence

© 2024 VCPOST.com All rights reserved. Do not reproduce without permission.

Join the Conversation

Real Time Analytics