"Chasing OpenAI"... Meta Focuses Intensely on AI Technology Development

WSJ "Zuckerberg Forms Separate Team Early This Year"
Busy Preparing for Data Center and Semiconductor Acquisition

Meta Platforms, the parent company of Facebook and Instagram, is accelerating the development of a new generative artificial intelligence (AI) system to catch up with OpenAI, the developer of ChatGPT, the Wall Street Journal (WSJ) reported on the 10th (local time).

WSJ, citing multiple sources, reported that Meta is focusing on building a powerful new AI system at the level of OpenAI's latest AI model GPT-4, which was released last March.

[Image source=AFP Yonhap News]

Meta is aiming to take the lead in the AI industry by developing 'Llama' technology, a large language model (LLM). It is reported that Meta plans to achieve this goal by upgrading 'Llama 2,' which was released in July, by next year.

To this end, Mark Zuckerberg, Meta's CEO, formed a separate team earlier this year to accelerate the development of generative AI tools. Meta is reportedly planning to start training the new LLM early next year. It is also expanding data centers necessary for AI training and securing Nvidia's AI semiconductor H100.

Sources revealed that Meta has also begun preparations for training the new AI model in partnership with Microsoft's (MS) cloud computing platform Azure.

WSJ stated, "The plans related to the new model are part of CEO Zuckerberg's efforts to widen the gap with competitors and take the lead in the AI world," adding, "CEO Zuckerberg is pushing to release the new model as open source, like Meta's initial AI products, so that many companies can use Meta's technology for free."

Since first unveiling 'Llama' in February, Meta has been aiming to dominate the AI market through an open-source strategy. This approach differentiates Meta from companies like OpenAI, MS, and Google, which have kept their LLM technologies confidential. When Llama 2 was released, it attracted significant attention by allowing free use even for commercial purposes.

Meta's open-source strategy helps to widely disseminate the technology, but some concerns have been raised that, given the still insufficient AI regulations and the many challenges inherent in the technology itself, problems such as the spread of misinformation could occur.

This content was produced with the assistance of AI translation services.