Meta bets on AI big time, according to Zuckerberg Meta bets on AI big time, according to Zuckerberg

What is “artificial general intelligence,” also known as AGI? Apparently, nothing less than the holy grail of the AI race. Wikipedia gives two definitions of AGI, one pronouncing it an intelligent agent that, if realized, “could learn to accomplish any intellectual task that human beings or animals can perform,” and another calling it “an autonomous system that surpasses human capabilities in the majority of economically valuable tasks.” Both definitions suggest limitless applications of AGI, and, when you put your alarmist hat on, inherent danger to us the humans. Pursuing the former and largely ignoring the latter, key companies in this domain – OpenAI, DeepMind, Anthropic, etc., – push on with their programs to make artificial general intelligence a reality. Meta, the company behind Facebook, has recently announced its plan to outshine everybody else in this field.

Meta and its artificial general intelligence development effort

Zuck at Threads, talking about AGI. Snapshot from ThreadsZuck at Threads, talking about AGI. Snapshot from Threads

Last Thursday, Mark Zuckerberg posted a two-minute Threads video dedicated to Meta’s AGI plans. Specifically, he said that 

  1. his company will stick with the policy of making its AI projects open source, 
  2. there are plans to closely integrate AI and Metaverse,
  3. the infrastructure supporting the said projects will be ramped up significantly,
  4. the best way to use AI and communicate therewith is through AR/VR goggles.

(We’ve recently posted a comparison of Apple’s Vision Pro and its competition, check it out: 4 alternatives to Apple's Vision Pro)

From the technical point of view, the big news is item #3 in the list above, plus the articulated fact that Meta is currently training LLama 3, its flagship AI. This means faster progress along the way to artificial general intelligence, capitalization on intermediate achievements in the process, and a signal that Zuckerberg bets big on AI, being pretty confident he can win the race or at least not lose it.

Should we be anticipating or worried?

While AI uprising is a popular sci-fi horror tale, and there is a substantial backing to the campaign to limit, or even halt the development of artificial general intelligence, the upsides of bringing one to life, so to say, balance the risks or even outweigh them. Imagine outsourcing mundane, routine, boring tasks to an entity that will complete them quickly and flawlessly, or myriads of calculations enabling a scientific breakthrough that, without AI, would have taken years. The prospects are promising. And, if AGI will really be intelligent, it’s likely to take a hush-hush path to domination, not rain all nukes on our heads. Hopefully.

Author's other posts

New rules for EU’s App Store: impact for developers and end users
Article
New rules for EU’s App Store: impact for developers and end users
Apple dodges a €500M EU penalty by overhauling App Store rules. EU users gain freedom with links to external payments, but face security risks; global changes are unlikely.
AI-based services for all: AI chatbots and platforms for children
Article
AI-based services for all: AI chatbots and platforms for children
AI is everywhere, in the kindergartens and primary schools. It's just the way things are. Here is a list of AIs and learning platforms that are child-safe.
OpenAI and LoveFrom’s secret product is an audio computer?
Article
OpenAI and LoveFrom’s secret product is an audio computer?
OpenAI and LoveFrom face a lawsuit by IYO over 'IO' branding. The battle heats up as both work on revolutionary voice-based AI devices, teasing a game-changing future.
Family Safety blocks Google Chrome in Windows: background and fixes
Article
Family Safety blocks Google Chrome in Windows: background and fixes
A filtering bug in Windows 11's Family Safety stops Google Chrome from launching. Here are a couple workarounds.