The company will create a new product group focused on Generative AI, a new machine learning technique that enables computers to generate text, draw, and make other media that mimic human performance.
The move comes when big tech companies and well-capitalized startups are touting advances in machine learning techniques and incorporating artificial intelligence models into their products.
The division will merge several teams into Meta, CEO Mark Zuckerberg said in a Facebook post. This new group will be managed under the Chief Product Officer, Chris Cox.
Zuckerberg said the team would develop “creative and expressive” tools for the Meta product.
“We’re testing experiences with text (such as WhatsApp and Messenger chats), with images (such as Instagram’s creative filters and ad formats), and with video and multimodal experiences,” Zuckerberg said. “We still have a lot of ground work to do before we get to a truly futuristic experience, but I’m excited for all the new things we’ll be building along the way.”
For example, the large language model built by OpenAI has been integrated into Microsoft’s Bing chatbot and into a separate chatbot called ChatGPT. Google
also working on a chatbot called Bard.
On Monday, Snap announced it would integrate its ChatGPT bot into its Snapchat app.
Last week Meta announced its new big language model called LLaMA. Meta said at the time that their models were popular because they were accessible to researchers and were smaller and cheaper to work with than larger models.
However, this announcement is also a sign that Meta, who has done a lot of research in artificial intelligence, will not let its competitors qualify in the AI race.