Meta plans to make advanced AI tools 'widely available' to the public despite Mark Zuckerberg's warning that they could ultimately pose a threat to humanity Artificial intelligence experts are now on high alert.
on video posted on his Facebook accountZuckerberg said building the “next generation of services” will require the development and widespread release of artificial “general intelligence” (usually meaning AI systems with human-level cognitive capabilities). Ta.
“This technology is so important and the opportunity is so great that it should be responsibly open sourced and made as widely available as possible so everyone can benefit from it,” Zuckerberg said. said.
However, Wendy Hall, a prominent UK-based computer scientist and member of the United Nations' AI advisory committee, says: told the Guardian He said Zuckerberg's plan was “really very scary” given the potential risk of abuse.
“It’s really scary to think that open source AGI will be released before we have a way to regulate these very powerful AI systems,” Hall said. “If technology like this were to fall into the wrong hands, it could do tremendous harm. It's so irresponsible for companies to propose it.”
“Open source” refers to the concept of making the underlying source code of a product available to everyone to see and use as they see fit.
Some experts are wary of open-sourcing AI, saying open-sourced AI tools exacerbate risks such as the spread of misinformation, election interference, job losses, and even a loss of human control over society. He claims to do so.
Elon Musk and former Google boss Eric Schmidt have warned that advanced AI could pose a world-ending risk if proper safeguards are not put in place.
Hall noted that true advanced general intelligence is “still many years away”, giving governments time to develop appropriate regulations for the burgeoning technology.
Andrew Rogowski, another AI expert at the University of Surrey in the UK, argued that regulators, not metas, should decide whether open source is safe.
“There are deep and complex debates about the merits of open sourcing current AI models, and pushing them into the realm of AGI could be world-saving or catastrophic.” Rogoiski told the media. These decisions need to be made by international consensus, not in the boardrooms of tech giants. ”
The Post has reached out to Mehta for comment.
Zuckerberg appears to have been hedging his bets during the trade Another interview with The Verge On Meta's AI ambitions – he told the tech site that he has not yet made a final decision on whether to open source advanced AI.
“I think we generally want to lean toward open source as long as it makes sense and it's a safe and responsible thing to do,” Zuckerberg said. “Of course, you don't want to be bound to doing something just because you say you're going to do it.”
Last year, Meta released an open source version of Llama 2, a large-scale language AI model. Zuckerberg said that as part of Meta's plan to accelerate innovation, he has ordered the company's two main AI divisions, FAIR and GenAI, to work more closely together.
Zuckerberg also argued that “new advice” will be needed to enhance the public's experience with advanced AI assistants.
“Over time, many of us will be interacting with AI frequently throughout the day,” Zuckerberg said. “I think many of us will be doing that using glasses, which are the ideal form factor to tell AI to see what we see and hear what we hear. They are always available to help.”
Meta is in fierce competition to develop advanced AI tools from rivals such as Google and Microsoft-backed OpenAI.
