Meta is testing a new feature where its AI chatbot will send "follow-up messages" based on past conversations to increase user engagement. According to Business Insider, Meta AI sends the first message without any prompt from the user and restarts the chat based on previous topics discussed.
This new Meta feature will be implemented on platforms like Messenger, Instagram, and WhatsApp.
An example chatbot used in tests, called "The Maestro of Movie Magic," directly messages users with questions like "Would you like a movie recommendation for tonight?"
It re-engages users based on conversations they started earlier.
This system is being developed according to guidelines prepared by the data labeling company Alignerr.
Meta and Alignerr are working together on training customizable AI bots that will proactively communicate with users in the future.
A Meta spokesperson told TechCrunch that this feature is being tested to "spontaneously" reconnect with users and aims to increase user engagement.
Meta AI can only send messages if the user has interacted with the chatbot at least once in the past 14 days.
Additionally, there must be at least a 5-message conversation history.
If the user does not respond to the first follow-up message, the bot will not send a second message.
Through Meta AI Studio, users can create AI bots tailored to their own experiences. These bots could eventually become systems that message users proactively.
Meta AI currently performs many functions via Facebook, Messenger, and Instagram, including chatting, sticker creation, and photo editing.
It was integrated into WhatsApp earlier this year.
Also, users can share their AI conversations in the "Discover" tab, but this can lead to unintentional exposure of personal data by unaware users.
Meta's new feature may raise concerns about user privacy and data security. Although spam prevention mechanisms are defined, AI-powered unexpected messages could be unsettling for some users.
