WhatsApp looks to be developing a new generative AI function that would allow users to create bespoke avatars of themselves for usage in any imagined environment. The in-development feature, which WABetaInfo discovered in the current WhatsApp Beta for Android 2.24.14.7, appears to generate photos using a combination of user-supplied photographs, text prompts, and Meta’s AI Llama model.
WABetaInfo shared a screenshot in which users can envision themselves “in any setting from the forest to outer space.” The examples appear to be quite typical of those produced by AI generators, especially if you’ve used applications like Lensa AI or Snapchat’s “Dreams” selfie function.
WhatsApp users will need to “take photos of yourself once” to generate the personalized avatar, which will then be used to train Meta AI to make images in the user’s likeness in any situation. WhatsApp users can then create avatars by saying “Imagine me” with a description of the setting in Meta AI chat or “@Meta AI imagine me…” in other WhatsApp conversations.
The function will apparently be optional, requiring users to actively enable it in their WhatsApp settings before using it. WABetaInfo also states that the reference photos can be erased at any moment using the Meta AI settings.
There is no indication of when the new Imagine AI selfie function will become widely available. WhatsApp is still rolling out support for its Meta AI chatbot, as well as more broad real-time AI image generation for US users, so the new AI avatar feature may not be available right away. Given the challenges with its earlier generative AI tools, Meta is justified in taking a more cautious approach.
