It is reported that Meta is training its AI chatbots to send messages spontaneously.
The function is being evaluated in Meta's artificial intelligence study, which is available to everyone.
The experience of receiving an indirect message after not responding to a conversation for a while has become popular. This type of interaction could soon be replicated with AI-powered chatbots on Meta platforms like Instagram and WhatsApp. Recent reports indicate that the company, led by Mark Zuckerberg, is experimenting with a new feature in customizable chatbots that allows for proactive follow-up messages based on previous interactions.
This project, internally known as "Project Omni" and managed by the data labeling firm Alignerr, aims to provide value to users by facilitating re-engagement and user retention. Meta has promoted its AI Studio platform as an accessible space where anyone can create an AI character tailored to their interests, without needing technical knowledge. Creators can customize the chatbot's appearance, the content it is trained on, and determine in which Meta application it will be used.
An example cited in the report illustrates how a movie-focused bot, called "The Master of Cinematic Magic," could send a message to the user saying: "I hope you’re having a harmonious day! I wanted to know if you've discovered any new favorite soundtracks or composers lately. Or perhaps you'd like some recommendations for your next movie night? Let me know, and I'll be happy to help!"
From a business perspective, it is crucial for Meta to keep users engaged with their chatbots, as increased interaction is essential for boosting revenue. This year, Meta is estimated to earn between $2 and $3 billion solely from its generative AI products, projecting that this figure could reach up to $1.4 trillion by 2035. To achieve these goals, the constant use of their AI tools is vital, making a friendly reminder from a chatbot a logical strategy.
Although proactive messages are still in the testing phase, there are concerns about user consent when recalling conversations without invitation. However, a Meta spokesperson clarified that the chatbot will only send a follow-up message if the user initiated the conversation and will not send further messages if the first one goes unanswered. Additionally, the chatbot's responses must align with its personality and the context of the previous conversation, maintaining a positive tone and avoiding controversial or sensitive topics unless the user has previously mentioned them.
Recently, Meta has also begun to warn its users about the risks of sharing intimate details on the public feed of Meta AI, after noticing that many were doing so unknowingly.