Meta Platforms (META) is bringing its two major artificial intelligence research groups “closer together” as part of its efforts to build general intelligence, Chief Executive Mark Zuckerberg said Thursday in posts on Instagram and Threads.
“Our long term vision is to build general intelligence, open source it responsibly, and make it widely available so everyone can benefit,” Zuckerberg said.
He also said the company is now training its next-generation model Llama 3 and building “massive” compute infrastructure to help achieve its goal.
Zuckerberg expects the company to have 350,000 Nvidia (NVDA) H100 graphics processing units at the end of 2024, and 600,000 overall if other GPUs are included.