Whatsapp is working on developing a feature that will allow users to generate images of oneself using Meta AI and also a function that allows you to choose the model Llama with which to perform functions based on artificial intelligence. This is reported by WABetaInfo – a portal that specifically deals with finding and making known the new features regarding the famous instant messaging app – which has identified the two new functions in the WhatsApp beta version for Androidspecifically in the updates marked with versions 2.24.14.13 And 2.24.14.7. It is currently unclear when these changes will arrive in the public version of WhatsApp.
As for the WhatsApp beta update for Android that brings the version number 2.24.14.7, WABetaInfo has identified a feature through which users will be able to choose between different Llama models for their interactions with the AI. As seen in the screenshot below, specifically it will be possible to opt for the model Llama 3-70B (which is set as the default option) if you need to make simple and quick requests or, alternatively, you can select the advanced model Llama 3-405B to receive output to more complex requests.
It is worth noting that, as you can see in the template selection menu interface in the WhatsApp beta, the Llama 3-405B template will be available for a limited number of prompts each week, and once you reach that limit, you will be able to continue the conversation with the AI using the default template.
Regarding the version 2.24.14.13 of WhatsApp beta for Android, instead, this brings a novelty related to the generation of images of users based on artificial intelligence. In particular, this new function should allow you to generate new artificial images of yourself after having provided some real ones to feed to the generative model based on AI integrated in WhatsApp. As underlined WABetaInfo «Users will retain full control over this feature, being able to delete preset photos at any time through Meta AI settings.»
Once users have taken photos of themselves and made them available to Meta AI, they can ask it to generate an artificial image of themselves by typing “imagine me” in the conversation with the chatbot. In addition to this, it will be possible to invoke the function in other chats by typing “@Meta AI imagine me”.
It is important to note that, given the implications related to privacy and security, the function is not active by default (it must be manually enabled in the settings) and, in any case, it can be deactivated at any time.
