Meta is bringing the voices of Judi Dench, John Cena and Keegan-Michael Key to its AI chatbot

New York CNN  —  Facebook and Instagram users will now be able to talk to voices that sound a lot like John Cena and Judy Dench. They won’t be the real actors, though, but an artificial intelligence chatbot. Parent company Meta (META) announced Wednesday that it is adding voice conversations and celebrity voices to its artificial intelligence chatbot, Meta AI. Now, instead of simply messaging with the chatbot, users can have real-time conversations and can choose from a selection of computer-generated or celebrity voices. The company partnered with Cena and Dench, as well as actors Kristen Bell, Awkwafina and Keegan-Michael Key, to train the chatbot to replicate their voices. The update comes as Meta seeks to help its AI chatbot — which users can chat with on Facebook, Instagram, WhatsApp and Threads — keep pace with competitors’ products, including ChatGPT, which is rolling out its own advanced voice mode. Meta CEO Mark Zuckerberg said Meta AI is on track to be “the most used AI assistant in the world” by the end of this year, likely helped by the more than 3 billion people who use the company’s apps each day, although it’s not clear how Meta is measuring use of the chatbot and how frequently people engage with the tool. Rival OpenAI came under fire earlier this year when it showed off its own real-time voice mode feature for ChatGPT because of a demo voice that sounded remarkably like actor Scarlett Johansson, who said she had been asked to participate in the project but declined. OpenAI denied that the voice, dubbed Sky, was based on Johansson, but paused its use anyway. Unlike that debacle, Meta appears to have formed formal partnerships with the actors whose voices were used to train its tool. LinkedIn on App Store displayed on a phone screen is seen in this illustration photo on June 5, 2024. Jakub Porzycki/NurPhoto/Getty Images Related article Social media platforms are using what you create for artificial intelligence. Here’s how to opt out Zuckerberg announced the new voice mode during his keynote speech at the company’s annual Meta Connect conference, where he also shared other AI advancements, a new, cheaper version of Meta’s Quest headsets and updates to the company’s line of augmented reality RayBan glasses. Among the other notable announcements: Meta will now let social media influencers make AI versions of themselves. Previously, influencers could train AI to have text conversations with their followers; now, followers will be able to have full, quasi-video calls with the AI versions of influencers who use the tool. Meta’s AI technology will also auto-translate and dub foreign language Reels (Meta’s short-form videos) for viewers. So, if you speak English but a Reel comes across your feed that was originally created in, say, Spanish, it will appear in your feed as if it were made in English, complete with edits to the speaker’s mouth to make the dubbing look natural. Meta’s AR glasses are getting live, AI-enabled translation, too. A user can have a conversation with someone speaking in a foreign language and, within seconds, hear the translation to their own language in their ear, Zuckerberg said.