Mark Zuckerberg has faced accusations of being irresponsible in his approach to artificial intelligence after working to develop AI systems as powerful as human intelligence. The Facebook founder has also raised the possibility of making it available to the public for free.
Meta’s CEO announced that the company intends to build an artificial general intelligence (AGI) system and plans to open source it, making it accessible to outside developers. He emphasized that the system should be “responsibly made as widely available as possible.”
In a Facebook post, Zuckerberg stated that the next generation of technology services requires the creation of complete general-purpose intelligence.
Although the term AGI is not strictly defined, it generally refers to a theoretical AI system capable of performing a range of tasks at a level of intelligence equal to or exceeding that of humans. The potential emergence of AGI has raised concerns among experts and politicians worldwide that such a system, or a combination of multiple AGI systems, could evade human control and pose a threat to humanity.
Zuckerberg expressed that Meta would consider open sourcing its AGI or making it freely available for developers and the public to use and adapt, similar to the company’s Llama 2 AI model.
Dame Wendy Hall, a professor of computer science at the University of Southampton and a member of the United Nations advisory body on AI, expressed concern about the potential for open source AGI, calling it “really, very scary” and labeling Zuckerberg’s approach as irresponsible.
According to Mr. Hall, “Thankfully, I think it will still be many years before those aspirations become a reality.” She stressed the need to establish a regulatory system for AGI to ensure public safety.
Last year, Meta participated in the Global AI Safety Summit in the UK and committed to help governments scrutinize artificial intelligence tools before and after their release.
Another UK-based expert emphasized that decisions about open sourcing AGI systems should not be made by technology companies alone but should involve international consensus.
In an interview with tech news website The Verge, Zuckerberg indicated that Meta would lean toward open sourcing AGI as long as it is safe and responsible.
Meta’s decision to open source Llama 2 last year drew criticism, with some experts likening it to “giving people a template to build a nuclear bomb.”
OpenAI, the developer of ChatGPT, defines AGI as “an AI system that is generally smarter than humans.” Meanwhile, Google DeepMind’s head, Demis Hassabis, suggested that AGI may be further out than some predict.
OpenAI CEO Sam Altman warned at the World Economic Forum in Davos, Switzerland, that further advances in AI will be impossible without energy supply breakthroughs, such as nuclear fusion.
Zuckerberg pointed out that Meta has built an “absolutely huge amount of infrastructure” to develop the new AI system, but did not specify the development timeline. He also mentioned that a sequel to Rama 2 is in the works.
Source: www.theguardian.com
