- By Ashish Singh
- Mon, 11 Sep 2023 12:02 PM (IST)
- Source:JND
Meta, formerly Facebook, is preparing to build a new artificial intelligence (AI) model that will compete with the power of OpenAI's latest chatbot, GPT-4, says a report. Meta is reportedly investing heavily in its AI capabilities, including the acquisition of additional Nvidia H100 AI-training chips and the expansion of its infrastructure.
Furthermore, according to The Wall Street Journal, Meta intends to begin training for this new large language model (LLM) in early 2024. Meta's CEO Mark Zuckerberg, is reportedly likely to make the new AI model available to companies and researchers for free at first. This strategy seems to be consistent with Meta's objective of ‘democratising AI’ and making it freely available for a variety of uses.
READ: Meta Threads Now Rolling Out 'Keyword Search' Feature To More Countries; Is India On The List?
Meta has also created a specialised team to develop this AI model, with a specific emphasis on developing AI tools that can effectively mimic human expressions, the report added.
Previously, Meta developed Code Llama, an AI tool aimed at producing new code and debugging human-written code. Code Llama is a large language model (LLM) that can produce and discuss code using text prompts. Its goal is to streamline developer operations and make the learning process easier for aspiring coders.
Code Llama, a specialist version of Llama 2, has the potential to be a productivity tool for developers as well as an educational resource for people wishing to improve their coding skills.
CEO Mark Zuckerberg announced earlier this year the development of a specialised product team focused on generative artificial intelligence (AI), showing Meta's interest in AI and the intentions to soon introduce something innovative.
Meanwhile, Meta is all set to bring its Quest 3 to the global market soon. With this, the company has already started bringing down the prices of Quest 2 to make a good market.
