Meta aims to match OpenAI with its upcoming GPT-4 competitor

Meta, once known as Facebook, is in the midst of creating an AI model that could stand shoulder to shoulder with OpenAI's GPT-4, as revealed by The Wall Street Journal.

This strategic move by the tech titan underscores its intent to match strides with other industry heavyweights in the rapidly evolving AI landscape.

Insiders hint that Meta's forthcoming AI offering might surpass the capabilities of its current Llama 2 AI. The model, anticipated for a release next year, is being designed as a robust tool for businesses, especially in areas of text generation and analysis. This direction seems to echo the trajectory of OpenAI's ChatGPT enterprise, which has been in the limelight since its recent introduction.

In a notable twist, Mark Zuckerberg, the helm of Meta, is believed to be championing the idea of making this new AI model open source. While the open-source approach promises adaptability and broad access, it's not without potential pitfalls. The flexibility of such models has raised eyebrows, with concerns about their misuse in spreading misinformation or potential copyright breaches. Legal teams at Meta have voiced apprehensions about the challenges tied to an open-source stance.

Zuckerberg has consistently highlighted AI as a pivotal area for the company's future, a sentiment echoed in recent earnings discussions. Yet, the path hasn't been entirely smooth. Documents from the previous year, shared by Reuters, suggest that Meta recognized its lag in AI research and investments when compared to giants like OpenAI and Microsoft.

To counteract this, Meta has been on a spree, acquiring H100 chips, which are crucial for training advanced AI models. Industry watchers anticipate that Meta's acquisition pace might position it as the world's second-largest holder of H100 chips by next year.

Meta's decision to lean into open sourcing appears to be a calculated attempt to diminish the lead of its competitors. This was evident when the company declared its Llama 2 model open source earlier this year, offering it freely for research and commercial endeavors.

As the industry awaits more details on Meta's AI ambitions, the company remains tight-lipped, not offering any official statements on the recent rumors.