In a new development, Meta, formerly known as Facebook, has introduced a series of celebrity chatbots as part of its AI technology. These chatbots are currently available in beta on Facebook Messenger, Instagram, and WhatsApp. The chatbots are modeled after famous figures such as Paris Hilton, Snoop Dogg, and Mr. Beast. For instance, there is a chatbot based on Snoop Dogg as a dungeon master and another based on Mr. Beast, which Meta describes as “the big brother that roasts you—because he cares.” In the future, Meta plans to make the tools used to build these celebrity chatbots available to its users and businesses.
One notable aspect that sets Meta apart from its competitors is its approach to building products based on open-source machine learning models. Google and OpenAI, for example, keep their latest AI models proprietary. Meta, on the other hand, made its Llama models open to the public in February and released Llama 2 in July. The models have been downloaded 30 million times, with approximately 7,000 derivatives created. This strategy allows outsiders to adapt Meta’s open-source AI code, which in turn helps inform how the company uses the project for its own apps and services.
While Meta’s open-source AI strategy has been praised for enabling faster innovation, it has also faced criticism. Some activists, like Holly Elmore, argue that Meta should stop distributing the most detailed versions of the Llama model due to concerns about AI safety. Elmore plans to hold a protest outside Meta offices in San Francisco this week. Despite these controversies, today’s launch of Meta AI is not the company’s first foray into AI technology. Previously, under the name Facebook, the company introduced a virtual assistant named M in 2015 to compete with Alexa and Google Assistant, but it was quietly shut down in 2018.