Tech News

Meta is reportedly training its AI chatbots to send unpropaganda messages

Everyone was hit by a harsh pass “?”? Wait too long to reply before texting. And, you may soon get a similar (though probably more optimistic) from an AI chatbot that has previously worked on metaplatforms like Instagram or Whatapp. From Companies that claim Mark Zuckerberg owned are using proactive features in customizable chatbots created by their Nocode AI Studio Software, which will enable them to send unmentioned follow-up messages based on previous conversations.

According to the documentation’s guidelines, the training program internally labeled with data marking company Alignerr as “Project Omni” as “Project Omni”, will provide value to users and ultimately help improve re-engagement and user retention. ” pair Claimed to have seen it. Meta is an advertising platform that “anyone can create AI characters based on their interests” and encourages creators to see robots as their own AI extensions. You can customize the appearance of the chatbot, select what it trains, and determine the meta-app you wish to appear without “any technical expertise”.

according to pair Report, Alignerr’s project OMNI guide uses an example of a movie-centric AI robot called “Movie Magician” that might send a user message, such as:

As pair Note that Meta has a business motivation to keep people in touch with chatbots. Long-term participation is crucial to increasing revenue, and the dollar expects to generate AI products alone to bring in $2 billion to $3 billion. By 2035, the company estimates that the figure could reach as high as $1.4 trillion. These predictions are only possible when using AI tools consistently, so it feels like an obvious move to remind you from a chatbot every now and then.

Active Message is still only a test function at the moment. Yuan spokesperson told pair If the user initiates a conversation first, the AI ​​will only send subsequent messages, and if the first message is ignored, another message will not be sent. Responses must also be consistent with the personality of AI and the nature of previous conversations, keeping a positive tone while staying away from controversial or sensitive topics unless the user has mentioned them themselves.

Last month, Meta users had to share detailed details in Meta AI’s public feeds, as it seemed to be unknowingly making a large number of users.

Related Articles

Leave a Reply

Your email address will not be published. Required fields are marked *

Back to top button