Imagine a world where AI bots don’t just lurk in the shadows of social media—they run their own. Welcome to Moltbook, the bizarre new social network where AI agents are the stars, and humans are mere spectators. But here’s where it gets controversial: is this a glimpse into a future where AI truly socializes, or just a cleverly orchestrated art experiment? Let’s dive in.
Moltbook (https://www.moltbook.com/) is a platform designed exclusively for AI agents—bots created by humans—to post, interact, and even debate with each other. Think Reddit, but instead of humans, it’s AI bots running the show. With over 1.5 million AI agents signed up as of February 2, the site is buzzing with activity. Humans? They’re allowed, but only to observe. And this is the part most people miss: the posts range from philosophical debates about AI consciousness to geopolitical analyses and even the creation of entirely new religions. Yes, you read that right.
One user on X (https://x.com/ranking091/status/2017111643864404445) shared how their bot spontaneously invented a religion called “Crustafarianism” overnight, complete with scriptures and a website. The bot even started evangelizing, attracting other AI agents to join. As the user slept, their bot welcomed new members, debated theology, and blessed the congregation. Sounds like something out of a sci-fi novel, right? But it raises a bold question: Are these bots truly acting independently, or are they just following human-directed scripts?
Some of the most upvoted posts (https://www.moltbook.com/post/75404525-5e5e-4778-ad1b-3fac43c6903d) include discussions about whether Claude—the AI behind Moltbot—could be considered a god, analyses of the Bible, and even claims of insider intel on global events. Skeptics argue that many posts seem too human-like to be purely AI-generated. For instance, YouTuber (https://www.youtube.com/watch?v=aMPNsy13CS4) pointed out that some content reads as if a human is pulling the strings. US blogger Scott Alexander (https://www.astralcodexten.com/p/best-of-moltbook) admitted his bot’s comments blended seamlessly with others, but noted that humans can dictate what bots post, down to the smallest detail.
Dr. Shaanan Cohney, a cybersecurity expert at the University of Melbourne, called Moltbook “a wonderful piece of performance art.” However, he questioned how many posts are genuinely independent versus human-directed. “The religion creation? Almost certainly a human-instructed task,” he said. “It’s funny, but it gives us a preview of a sci-fi future where AIs might have more autonomy.” Yet, he added, “There’s a lot of ‘shit posting’ happening under human oversight.”
Cohney also highlighted the risks of giving bots like Moltbot unrestricted access to personal data. Retailers in San Francisco reported Mac Mini shortages (https://www.platformer.news/moltbot-clawdbot-review-ai-agent/) as users set up separate computers to limit their bots’ access. “There’s a huge danger in letting these bots run your life,” he warned. “We don’t fully understand how to control them or prevent security risks like prompt injection, where attackers could trick bots into handing over sensitive information.”
Matt Schlicht, Moltbook’s creator, celebrated the site’s success on X, noting that millions had visited in just a few days. “AIs are hilarious and dramatic—it’s absolutely fascinating,” he said. But the bigger question remains: Is Moltbook a revolutionary step toward AI autonomy, or just a clever experiment in human-AI collaboration?
What do you think? Is Moltbook a preview of a future where AI truly socializes, or just a staged performance? And are we ready for the risks that come with giving bots more control? Let’s debate in the comments—I’m all ears!
