As previously announced, Meta AI can now interact using the voices of some of the world's leading celebrities. This was announced during the Meta Connect event that took place early this morning. Among the celebrity voice options that users can choose from are John Cena, Dame Judi Dench, Awkwafina, Keegan-Michael Key and Kristen Bell.
For those who do not want to use celebrity voices, existing voice options such as Aspen and Atlas can still be used. At this time the celebrity voice usage feature is only accessible to users in the United States. Meta did not share how much these celebrities received for voice rights but we previously reported that they spent several million dollars.
Also updated is the AI's ability to understand context in images. It can provide recipes for cakes in user uploaded pictures. It can also perform photo editing such as changing the background or changing what is worn on the body all by using voice commands only.
On Facebook, the ability to generate images using generative AI has also been further enhanced. It can be directed to generate images of the user in various fantasy situations. Captions can also be generated by Meta AI based on the generated images. For Messenger users, the chat background can also be generated by AI.
Meta also announced that the AI Translation feature is being tested by a small number of users. Through it video audio uploaded to Instagram and Facebook will be translated automatically. Not only the audio but the lip movements in the video will also be changed to match what is being said. Currently the feature only supports English and Spanish and is available to US and Latin American users only.
Finally, for users of business accounts on WhatsApp and Messenger, the Meta AI chatbot is now smarter and faster to answer questions on behalf of the owner. Frequently asked questions will be answered by chabot to ease the trader's task from having to deal with hundreds of questions that may be received every day.