According to a report from The Wall Street Journal, there are plans for the development of a new AI assistant by Meta, the parent company of Facebook and Instagram. The goal is to surpass the capabilities of the Llama 2 model, which was unveiled two months ago, and to compete with OpenAI’s GPT-4.
The training plan for this new model will commence in early 2024, as Meta will utilize its own data processing servers in its data centers, in contrast to Llama 2, which uses resources from Microsoft’s Azure. Currently, the company is also recruiting relevant employees and placing orders for a large quantity of NVIDIA’s H100 GPUs.
Meta will continue to adhere to its previous approach of open sourcing the AI model, allowing organizations to utilize and build upon it freely.
TLDR: Meta, the parent company of Facebook and Instagram, is planning to develop a new AI assistant with ambitions to surpass the capabilities of the Llama 2 model and compete with OpenAI’s GPT-4. The training of this new model will start in early 2024, utilizing Meta’s own data processing servers. Meta is also hiring employees and ordering a significant number of NVIDIA’s H100 GPUs. The company will continue its approach of open sourcing the AI model, enabling further application and development opportunities.