Meta's new AI feature: Create your own AI avatar on WhatsApp
Meta is testing a new feature that will allow users to create their AI-generated avatars on WhatsApp. Here is everything you need to know!
AI features are pouring from every corner now. When it comes to Meta's AI, it has often raised privacy concerns. Nevertheless, their new AI assistant is now available in India on WhatsApp. So far, Meta's new AI model can perform tasks like generating AI images, planning things, etc but it will soon help you create an AI avatar from your photo! Here is all you need to know about this AI feature coming on WhatsApp and how it will work!
Meta's upcoming AI feature
Wabetainfo recently teased a post that shows an interesting AI-powered feature that will be released soon by Meta. According to the tipster, WhatsApp is currently working on an optional feature with their latest AI model 'Llama'.
A sneak peek of this feature was also posted by the tipster claiming that Meta will soon allow users to create AI-generated avatars from setup images. However, this latest feature is still under development and was discovered it was being tested on the latest WhatsApp beta for Android 2.24.14.13 update.
Make an avatar with Meta AI: How this feature could look like
A screenshot of the possible feature unveils how this feature is going to work. To create an AI-generated avatar image of themselves, users need to take setup pictures. To produce an accurate AI representation of users, WhatsApp will ask you to click a selfie that will be analysed to generate the avatar. After that, users need to type a text prompt to Meta AI saying "Imagine Me" in the chat. Alternatively, users can also type "@Meta AI imagine me" in any chat to do the same.
Another AI privacy issue?
While AI has taken the spotlight for the right reasons, it has also been thrown under the bus due to data privacy issues. The main concern for AI users is that companies track, record, analyse and train AI models with their personal information. Microsoft recently received backlash for its "Recall feature" that took screenshots of a user at regular intervals.
Similarly, after Adobe updated its terms of service, it received heavy criticism from its customers on the usage of its data to train its AI model 'Firefly'. However, the firm later clarified that it does not store or use customer information for training AI systems.
Since Meta's avatar feature requires photos of users, many could worry about how this data would be handled. Even so, the tipster mentioned that this upcoming feature is safe and users will have complete control of their setup images which can be deleted in the AI setting.
When it comes to chat privacy, Wabetainfo explained that Meta cannot read the messages since the command is processed separately and the AI avatar image will automatically be shared in the chat.
Moreover, this feature will be optional and users can opt in or out according to their preferences.
The takeaway
Meta's AI will soon allow users to create AI avatar versions from their pictures but it will not roll out as a default feature amidst privacy concerns. While the firm ensures end-to-end encryption in the user's chat, your conversation with its AI is saved but anonymously.