According to what Meta trained on Chatbots from artificial intelligence to send unlimited messages

Everyone was hit with an urgent passage? Text after waiting for a very long time to respond. You may soon get a similar treatment (although it is likely to be more optimistic) than the AI ​​Chatbots that you previously participated in Meta platforms such as Instagram or Whatapp. New report from It claims that Mark Zuckerberg, owned by a proactive feature in customized Chatbots that were created using AI’s AI’s Studio program without a symbol, which will enable them to send unqualified follow -up messages based on previous conversations.

Internally known to the ALIRERRR data mark as the “OMNI Project”, the training project will provide “value for users and ultimately help improve the re -participation and retain user,” according to the guidelines mentioned in the documents Dualism He claims to have seen. Meta announces As a platform where “anyone can create the character of Amnesty International based on their interests” and encourages creators to look at robots as an extension of themselves Amnesty International. You can customize the appearance of Chatbot, choose the content that has been trained on and select the application of the description you want to appear, all without “any technical experience”.

According to Dualism Reports, the Alignerr omni project, is an example of AI BOT, which focuses on films called “The Maestro of Movie Magic” that might send a user message like: “I hope you will spend a harmonious day!

like Dualism Note, there is a Meta commercial incentive to keep people engaged with Chatbots. The lengthy participation is vital to increasing revenues, and Meta expects to bring $ 2 billion to $ 3 billion of AI obstetric products alone. By 2035, the company estimates that the number can reach $ 1.4 trillion. These types of predictions will not be possible unless artificial intelligence tools are constantly used, so a friendly remote from Chatbot from time to time seems to be a clear step.

Pre -proactive messages are still just a test feature. A Meta spokesman said it is certain that it seems to remember the talks and start new procedures without an invitation approaching a gray area for the user’s consent. Dualism The artificial intelligence will only send a follow -up message if the user has started the conversation for the first time, and will not send another message if the first message is ignored. Responses should also be consistent with the character of artificial intelligence and the nature of the previous conversation, while maintaining a positive tone while staying away from controversial or sensitive topics unless the user himself mentioned.

Last month, dead Its users did not exchange intimate details in the general summary of Meta AI after it appeared that a large number of users seemed unintentionally doing it.