The team behind the popular artificial intelligence (AI) chatbot ChatGPT has announced the formation of a team to mitigate and manage the risks of hyper-intelligent AI systems.
In an announcement on the OpenAI blog on July 5, the company said a new team would be created to “steer and control AI systems much smarter than us.”
In a statement, the nonprofit said it believes superintelligence is “the most impactful technology humankind has ever invented” and that, despite the risks, it can help solve many problems.
“The immense power of superintelligence can also be very dangerous, leading to the incapacitation of mankind, and even the extinction of mankind.”
OpenAI said it believes superintelligence could arrive within the next decade.
OpenAI said it aims to dedicate 20% of its already reserved computing power to the effort and create a “human-level” auto-alignment researcher. In theory, automated researchers would help the team manage the safety of superintelligence and align it with “human intent.”
The company has now appointed Ilya Sutskever, the company’s chief scientist, and Jan Rajke, the institute’s coordinator, as co-leaders of this effort. It solicited machine learning researchers and engineers to join the team.
Related: OpenAI Suspends ChatGPT’s Bing Feature After User Jumped Paywall
This announcement from OpenAI comes as governments around the world consider measures to control the development, deployment and use of AI systems.
European Union regulators are among the most advanced countries in introducing AI regulations. On June 14th, the European Parliament passed the first EU AI law. The bill would require systems like ChatGPT to disclose all AI-generated content and other measures.
The bill will need to be discussed in more detail in Congress before it takes effect. Nonetheless, the bill has sparked outcry from AI developers, who say it could stifle innovation.
In May, OpenAI CEO Sam Altman traveled to Brussels to discuss the potential negative consequences of overregulation with EU regulations.
U.S. lawmakers also introduced the National AI Commission Act, which would set up a commission to determine the nation’s approach to AI. US regulators have also outspokenly expressed a desire to regulate this technology.
On June 30, Senator Michael Bennett drafted a letter urging big tech companies, including OpenAI, to label AI-generated content.
magazine: BitCulture: Solana Fine Art, AI Music, Podcasts + Book Reviews