Meta has a new method to impose its AI and it is very intrusive

As a reminder, Meta offers assistants created from the AI ​​Studio platform which are able to send messages and memorize previous exchanges. Business Insider discovered this functionality via documents from a data labeling company that collaborates with Meta.

Among the examples, a personalized and passionate cinema AI spontaneously sends a user: “I hope you had a nice day. I wanted to get news and see if you discovered new soundtracks or composers recently. Or maybe you want recommendations for your next cinema evening.”

Meta confirms to Techcrunch that the company tests automatic recovery messages While Meta Ai is more popular than ever.

The functionality respects certain limits. Chatbots send reminders only within 14 days of a conversation started by the user and only if five minimum messages have been sent. Chatbots stop relaunching in the absence of an answer to the first message.

This technology looks like that offered by startups like Character AI and Relitika which allow their chatbots to launch the conversations themselves. Karandeep Anand, a new Character AI manager, joined the team last month after being vice-president of business products at Meta.

Meta specifies, however, that chatbots “are not professionals or experts trained to help people. Conversations with personalized AI cannot replace professional advice. You must not trust AI for medical, psychological, financial, legal or any other type of professional advice.”

In short, a philosophy that denotes that of Sam Altman, which is happy that users turn to AI for advice on their privacy.

Techcrunch Also asked Meta if an age limit was imposed for the recovery of the AI, which is not the case while the laws of Tennessee and Puerto Rico limit certain commitments with the adolescents.

Documents revealed last April that Meta provides for $ 2 to 3 billion in revenues via the generative AI in 2025 and possibly 1.4 Billion of dollars by 2035.

A projection that is based on income sharing agreements with companies that house these Open Source Llama models. Meta AI could possibly disseminate advertisements and offer a subscription.

Meta refused to comment on its IA marketing plans or the possible integration of advertisements or sponsored responses.

Want to save even more? Discover Our promo codes Selected for you.