Mark Zuckerberg Deems His New Artificial Intelligence the Most Powerful Yet

Mark Zuckerberg’s aims to develop artificial general intelligence (AGI) through maximizing the Nvidia GPU resources he acquired. He shifted his FAIR research group’s focus towards creating generative AI products to incorporate into his social media platforms. By the end of 2024, Meta plans to have a total of 600,000 Nvidia GPUs dedicated to AI work, making it one of the largest in scale for any company.

While the new version of Zuckerberg’s open-source AI, LLaMA 3, is considered advanced, it cannot run any proprietary code other than its own. Zuckerberg believes that LLaMA 3 poses no existential threat and that AI is not likely to create any fatal issues at this stage of development. Despite AI technology’s essential role in human life, it is not seen as a threat to humanity’s future decisions.

In a discussion with The Verge, Zuckerberg expressed concerns about multimodal AI aspects, particularly those capable of creating images or videos, along with text assistance. As the U.S. faces a challenging election year, he is apprehensive about the potential implications of multimodal AI, especially with image manipulation technologies.

Zuckerberg also hinted at Meta’s progress in developing consumer neural interfaces, wearable devices that interpret brain signals to control computers. Unlike Neuralink’s invasive brain chips by Elon Musk, Meta’s neural bracelet can be worn on the wrist and reads signals from the brain to control devices via nerve impulses. The bracelet uses electromyography (EMG) to translate brain signals into hand gestures that control devices, with potential consumer availability in the upcoming years. Although the technology is still in its early stages, Zuckerberg finds it fascinating and promising for future applications.