FILE – Facebook’s Meta logo sign is seen at the company headquarters in Menlo Park, Calif., on, Oct. 28, 2021. Officials at Meta say they have found and disabled a network of thousands of fake Facebook accounts linked to China that were used to spread partisan content in the U.S. The accounts disclosed on Nov. 30, 2023, were designed to look like they were run by everyday Americans. (AP Photo/Tony Avelar, File)
(NewsNation) — Meta CEO Mark Zuckerberg announced Thursday that the company is investing in the development of human-like artificial intelligence that will be widely available.
“Our long term vision is to build general intelligence, open source it responsibly, and make it widely available so everyone can benefit,” Zuckerberg said in an Instagram post.
The company is combining its two major AI research arms in order to build artificial general intelligence, or AGI.
“It’s become clearer that the next generation of services requires building full general intelligence,” Zuckerberg said. “Building the best AI assistants, AIs for creators, AIs for businesses and more.”
So, what is artificial general intelligence?
Generally speaking, it’s the idea that AI can think and reason like human beings. It’s defined by TechTarget.com as “the representation of generalized human cognitive abilities in software so that, faced with an unfamiliar task, the AGI system could find a solution.”
The definition is less clear for Zuckerberg. He told The Verge he has no “one-sentence, pithy definition.”
“You can quibble about if general intelligence is akin to human level intelligence, or is it like human-plus, or is it some far-future super intelligence. But to me, the important part is actually the breadth of it, which is that intelligence has all these different capabilities where you have to be able to reason and have intuition,” he said in an interview with the news outlet.
One thing is for certain: there’s a vast amount of infrastructure needed to achieve it.
Zuckerberg said on Instagram that by the end of 2024, the company will have around 350,000 of Nvidia’s H100 graphics processing units, the preferred chip for building AI. At a cost of around $30,000 for each one, the company is looking at $10.5 billion for the computing power.
The hardware is already being used to train Meta’s Llama 3 model, the company’s response to the tech behind OpenAI’s ChatGPT. The model won’t be kept locked away, but rather shared like Llama 2 was.
“This technology is so important and the opportunities are so great that we should open source and make it as widely available as we responsibly can,” Zuckerberg said.
The announcement follows the September launch of the Meta Ray-Ban smart glasses, which Zuckerberg indicated would be a major focus of the company moving forward.
“I think a lot of us are going to talk to AI frequently throughout the day, and I think a lot of us are going to do that using glasses,” Zuckerberg said. “Because glasses are the ideal form factor for letting an AI see what you see and hear what you hear, so it’s always available to help out.”
It also comes after Meta’s AI chief expressed skepticism that AGI is coming anytime soon.
Meta joins other tech companies like Google and Microsoft investing heavily in AI development. The increasing use of AI comes after OpenAI thrust the technology into the mainstream last year with its ChatGPT bot capable of quickly creating stories, memos, videos and drawings upon request.
The proliferation of the technology has raised concerns about risks including deepfake videos and other misinformation. Last year, tech CEOs met with White House officials and agreed to a set of safety guidelines.
The Associated Press contributed to this report.
