Zuckerberg: Meta Plans to Build Artificial General Intelligence

Facts

• Meta Chief Executive Officer Mark Zuckerberg announced Thursday that his company's goal is to create artificial general intelligence (AGI) — AI that equals or surpasses human intelligence. In a post to Meta's social platform Threads, he said the 'long-term vision is to build general intelligence, open source it responsibly, and make it widely available.'¹
• As part of the AGI pursuit, Meta will also shift its AI research group, called FAIR, to the same part of the company that contains its AI product development team. The race to AGI among tech companies has led to intense pursuits of talent, including offering compensation packages of over $1M. Zuckerberg also said Meta will own 340K of Nvidia’s H100 GPU computer chips — the top choice in the industry — by year's end.²
• Zuckerberg said he thinks many people will 'talk to AI frequently throughout the day,' particularly through the use of AI-incorporated glasses, because glasses let 'AI see what you see and hear what you hear.' Meta's AI technology has already been incorporated into RayBan eyeglasses.³
• While Zuckerberg has said that he doesn't think people will use AI for its ability to generate code, he said the company's new focus on AGI is based on the release of its large language model Llama 2. He believes the coding aspect is 'really important structurally for having the LLMs be able to understand the rigor and hierarchical structure of knowledge' and an 'intuitive sense of logic.'²
• Zuckerberg also argued that this doesn't mean the company is shying away from the Metaverse. He believes the Metaverse will become more popular as AI glasses grow in popularity.⁴
• In an interview with The Verge, he claimed Meta's AGI goals aim to make the technology open source, arguing that this method 'addresses a large class of issues that might come about from unequal access to opportunity and value.' Currently, the Llama language model can be considered open source.²

Sources: ¹CNN, ²Verge, ³India Today and ⁴Engadget.

Narratives

• Narrative A, as provided by Autoblocks Glossary. AGI isn't just another form of AI, but an existential threat to humanity as a whole. If AGI isn't vigorously monitored, it will eventually become smarter than humans and develop itself exponentially until we can no longer put the genie back in the bottle. Once it's fully autonomous, it can decide to overtake human beings and the environment we inhabit. This is more than just the latest social app, and it requires much more scrutiny.
• Narrative B, as provided by OpenAI. While there are risks involved in any advanced technology, we cannot forget about the incredible potential of AGI, particularly regarding the science. As AI companies slowly and cautiously roll out AGI to the world — and after periods of thorough safety testing — the world will be able to learn about and adapt alongside AGI to make their lives exponentially better economically and socially.

Predictions