Mark Zuckerberg’s Meta AI Feed Is the Saddest Place on the Internet

Mark Zuckerberg’s grand vision for AI has taken a strange, and at times, unsettling turn. With the launch of Meta AI’s stand-alone chatbot app earlier this year, the company introduced what it called a fun and engaging tool for interacting with artificial intelligence. But at the heart of this app is a public “Discover” feed — a stream of user interactions with the AI that feels less like innovation and more like accidental oversharing on a massive scale.

The Meta AI app allows users to ask questions, generate images, write poems, and engage in conversations with a chatbot. But unlike other popular AI platforms like ChatGPT or Google Gemini, Meta’s app includes a social element: a feed of conversations and prompts that users have chosen — or think they’ve chosen — to make public. While the feature was pitched as a way to “explore how others are using AI,” the result is a feed filled with raw, sometimes deeply personal content that often seems like it was never meant to be shared.

At first glance, the feed mostly appears harmless: AI-generated artwork, quirky requests for recipe suggestions, or a user asking for vacation tips. But dig a little deeper and it turns into something much more intimate — and uncomfortable. Users are openly asking the bot for legal advice, help writing letters to judges, health guidance, and even sharing emotional tributes to deceased loved ones.

Some of the most personal posts include full names, email addresses, and even phone numbers. In one case, a man asked the bot to help write a letter to the court for a child custody dispute. Another asked for a good morning prayer, his voice trembling with vulnerability in an audio message that was inadvertently shared publicly. It’s easy to imagine these people had no idea they were broadcasting their conversations to the internet.

Meta insists that chats aren’t public by default — users have to actively click “Share” and then “Post” to make a conversation appear in the Discover feed. A company spokesperson emphasized this multi-step process, suggesting users should be fully aware when they publish content. But given how many personal posts are showing up, it’s hard to believe everyone fully understands the consequences of clicking those buttons. The app doesn’t feel like a social network — it feels like a private space. That illusion may be exactly what’s making this situation so problematic.

What’s more eerie is that some shared posts include audio recordings. These clips reveal conversations that often don’t seem meant for the AI at all — one recording captures a discussion between coworkers about their shift schedules. Another sounds like a classic “pocket dial” where Meta AI pipes in suggestions while two people chat about something unrelated. It’s not just awkward; it’s invasive.

Adding to the strangeness is the fact that many posts display users’ actual Instagram or Facebook profiles. With just a few taps, you can often track down their real names and social media pages. It gives the whole feed the air of a surveillance window — less like browsing a tech feature and more like eavesdropping through someone’s diary.

Meta AI’s public feed might have been designed with good intentions — a space to showcase creativity and AI engagement. But in practice, it often resembles a confessional booth with the door left wide open. And instead of a priest, it’s Mark Zuckerberg’s algorithm listening in.

There are, of course, plenty of lighthearted moments scattered throughout. Some people are clearly using the AI for entertainment: generating funny images, composing poems, or brainstorming party ideas. But these glimpses of levity only underscore the dissonance of seeing deeply personal, sometimes painful moments published for all to see.

Zuckerberg has positioned Meta’s future around AI, with ambitious investments and goals to create a kind of digital superintelligence. But if the Discover feed is any indicator, the company still hasn’t figured out how to manage the human side of AI — the emotional messiness, the confusion, the very real risks of turning personal thoughts into public posts.

For now, Meta AI’s feed feels less like a bold future of human-AI collaboration and more like a cautionary tale: what happens when you hand over your secrets to a chatbot without realizing the whole world might be watching.