Mark Zuckerberg, CEO of Meta, announced on January 18th the company’s intensified efforts to integrate its business-focused artificial intelligence (AI) research with its Fundamental AI Research (FAIR) team to enhance technology productization. Zuckerberg highlighted the importance of AI in creating new generative models and products, mentioning significant investments in specialized computer chips.
Zuckerberg stated that the next generation of services requires full general intelligence, emphasizing the need for advancements in all AI domains to build top AI assistants. Meta plans to upgrade its technological infrastructure to include productive AI in its products, aiming to acquire around 350,000 H100 GPUs from chip designer Nvidia by year-end.
Mirroring Alphabet’s move to merge its two advanced AI research labs, Google Brain and DeepMind in 2023, Meta announced the unification of its two advanced AI research divisions. This strategy is part of an effort to catch up with Microsoft and OpenAI, who have made significant strides with AI chatbots like ChatGPT and the powerful language model GPT-4.
Meta has conducted significant AI research, from unsupervised learning, where AI learns patterns without labeled data, to creating AI software that outperforms top players in the strategy game Diplomacy. The company has also made progress in machine translation and computer vision algorithms.
Meta’s GenAI team developed Llama 2, a robust open-source language model. While not as advanced as OpenAI’s GPT-4 or Google’s Gemini models, Llama 2 is favored by developers for creating cost-effective and customizable chatbots compared to other models.