Meta is developing a next-generation artificial intelligence model to compete with OpenAI's GPT-4

Meta is investing in hardware and infrastructure to create a chatbot capable of competing with OpenAI's GPT-4 model.

Here's What We Know

According to The Wall Street Journal, the company plans to start training a large language model in early 2024. CEO Mark Zuckerberg insists it will be free for AI tool developers.

According to the report, Meta is actively buying Nvidia H100 chips to train neural networks and is beefing up its computing power so it doesn't have to depend on Microsoft's Azure cloud. Earlier this year, the company formed a team to accelerate the creation of AI that mimics human behaviour.

Meta's plans to create several models of generative AI were previously reported. It is rumoured to be testing a chatbot for Instagram with 30 "personalities". It may be referring to unannounced AI personas that will be unveiled this month.

Source: The Wall Street Journal