Why Mark Zuckerberg wants us to remove the stigma around virtual friends: ‘I’m not going to let this go’

“Here’s one stat from working on social media for a long time that I always think is crazy: the average American has fewer than three friends,” Mark Zuckerberg, CEO of Meta, the owner of Facebook, Instagram, and WhatsApp, said in a recent podcast. “And the average person has demand for meaningfully more. I think it’s something like 15 friends or something,” he added. Zuckerberg’s idea is that these extra friends will be robots, and that, of course, they will be provided by Meta AI.

Meta has just launched its standalone artificial intelligence (AI) app, Meta AI, in the U.S. In Europe, it’s only available as a pairing with Ray-Ban Meta smart glasses. Based on its Llama 4 model, Meta AI allows for conversations, image creation, and a proprietary audio format called super duplex that allows for fluid conversations, with interruptions and interludes (only available in English for the U.S., Canada, Australia, and New Zealand). In Europe, Meta AI is only available in its family of apps, like WhatsApp, not as a standalone app.

Along with hosting the first AI developer conference, LLamaCon, on April 29 to promote his open models, Zuckerberg toured several podcasts. In all his public appearances, he wore his Ray-Ban Meta glasses, which already include some AI. There, he explained how he believes there will be a place for all the big companies in the future of AI-powered apps: “There are going to be a lot of different companies doing cutting-edge work in different areas. Some will be more focused on the business world or programming. Others on productivity. And others, on social or entertainment.”

Meta clearly wants to focus on entertainment. Zuckerberg repeated the word “fun” more than a dozen times in a podcast. His efforts won’t be focused on competing for a work assistant with OpenAI, or for code with Cursor, but rather on the personal side, with memes and intimate conversations.

“Is this going to replace in-person connections or real-life connections? My default is that the answer to that is probably no,” he told Dwarkesh Patel’s podcast. “There are all these things that are better about physical connections when you can have them. But the reality is that people just don’t have the connections, and they feel more alone a lot of the time than they would like.”

Zuckerberg believes the “stigma” of AI for personal relationships will gradually disappear. “Over time, we’ll find the vocabulary as a society to articulate why they are valuable, why the people who are doing them are rational for doing it, and how it is actually adding value to their lives,” he added.

Conquering other apps

Meta wants to conquer a space already occupied by new apps like Character AI, Janitor, and Status, which offer different types of human relationships to their users and are hugely successful. Although its use is growing, the specific utility of AI in this sector remains to be seen. Meta AI also offers a tab to “discover” what other users are doing with AI, like a social network. Grok is increasingly prominent on X, where many users ask for context and clarification. OpenAI is also considering incorporating a social network into its ChatGPT to share AI creations.

“There are a handful of companies doing virtual therapists, virtual girlfriend-type stuff. But it’s very early. The embodiment in those things is still pretty weak. You open it up and it’s just an image of the therapist or the person you’re talking to. Sometimes there’s some very rough animation, but it’s not an embodiment,” said Zuckerberg, who hopes Meta AI will fill those gaps.

Within Meta, there are engineers who doubt Zuckerberg’s intentions. To make his AI more “fun,” he has to tolerate more potentially inappropriate content, including sexual content. In a test conducted by The Wall Street Journal with hundreds of test conversations, the bots, in their distinct voices, talked about sex, including with minors. Zuckerberg wants to eliminate caution and relax the limits on his bots. “Snapchat and TikTok got away from me, but this [referring to AI] isn’t going to get away from me,” Zuckerberg reportedly said, according to employees who heard his comments as reported by the WSJ.

The battle for the youth market is something that, as with social media, is already beginning to emerge in AI. Google has opened up the option for its Gemini chatbot to be available in the U.S. to children under 13, as long as it is controlled by their parents’ account and used for educational purposes.

The impact on users’ mental health in terms of potential sexualization or deep relationships with bots is, for Zuckerberg, a secondary issue: “There are a lot of questions that can only be answered when you start observing how people behave. Being too rigid at the beginning and saying, ‘We think this isn’t right,’ often prevents anything valuable from emerging,” Zuckerberg said. “People use things that bring them value. One of my key principles when designing products is that people are smart. They know what works for them in their lives. But if you think what someone is doing is wrong and that person thinks it’s bringing them something important, most of the time that person is right and you are wrong.”

Sign up for our weekly newsletter to get more English-language news coverage from EL PAÍS USA Edition