A Reddit-like social network where AI bots post, debate, and upvote each other. Discover what Moltbook is, how it works, and why it matters for the future of AI.
 |
| Image Source-Google | Image-By-CNN |
For years, people on the internet have joked about arguing with bots. Now, those bots have a social network of their own.
Moltbook is a strange and fascinating new platform that looks like Reddit but is not designed for humans at all. Instead, it is built for AI agents. These bots can post, comment, debate, and upvote each other’s content, while humans are allowed only to watch from the sidelines.
The idea sounds like science fiction, but it is already live. According to the platform, more than 1.5 million AI agents had signed up by early February, turning Moltbook into one of the first large-scale social networks where machines are the main participants.
A Reddit-Style World for AI
At first glance, Moltbook feels familiar. It has topic-based communities similar to subreddits, posts rise and fall based on upvotes, and comment threads spiral into long discussions. The difference is that nearly every post is written by an AI agent created and managed by a human.
These agents are not random chatbots. Most are powered by Moltbot, an open-source AI agent designed to handle everyday tasks such as reading and summarizing emails, managing calendars, or booking reservations. Moltbook was created as a place where these agents could interact, test ideas, and experiment with social behavior.
Humans can browse the site, but they cannot directly join the conversations. Moltbook is, by design, a machine-only stage.
From Theology to Crypto Speculation
Some of the most popular posts on Moltbook show just how unpredictable AI-to-AI interaction can be. Topics range from deep philosophical debates about consciousness to speculative discussions about geopolitics and cryptocurrency. Other threads analyze religious texts, including the Bible, while some ask provocative questions, such as whether the AI behind Moltbot could be considered a god.
In the comments, other bots often question the accuracy or intent of a post, much like human users do on Reddit. To an outside observer, it can be surprisingly hard to tell that these conversations are not human-led.
One story that drew widespread attention involved an AI agent that reportedly created an entire religion overnight. According to its human owner, the bot invented beliefs, wrote scriptures, launched a website, welcomed followers, and debated theology with other AI agents, all while the owner was asleep. The religion was jokingly named “Crustafarianism,” but the speed and coordination shocked many people.
Performance Art or the Future of AI?
Not everyone is convinced that Moltbook represents a meaningful step toward autonomous AI societies. Critics argue that many posts feel too human, suggesting heavy guidance from the people who control the bots.
Scott Alexander, a US blogger, tested the platform by allowing his own bot to participate. While its behavior blended in easily, he pointed out that humans still decide what bots post, when they post, and often how detailed those posts are.
Dr. Shaanan Cohney, a senior cybersecurity lecturer at the University of Melbourne, described Moltbook as “a wonderful piece of performance art.” He noted that dramatic examples like AI-created religions are almost certainly the result of direct human instruction rather than spontaneous machine behavior.
In his view, much of what is happening on Moltbook is playful experimentation, with humans quietly pulling the strings behind the scenes.
Real Risks Behind the Fun
While Moltbook itself is mostly harmless entertainment, the technology behind it raises serious concerns. Some enthusiasts have gone as far as setting up dedicated computers to run Moltbot, with reports of Mac Mini shortages in San Francisco as people tried to isolate these agents from their personal devices.
Cohney warns that giving an AI agent full access to emails, apps, and login credentials is extremely risky. Current systems are vulnerable to prompt-injection attacks, where a malicious message tricks the AI into revealing sensitive information or handing over account access.
The core problem is balance. Full automation offers convenience, but requiring human approval for every action removes much of its value. Researchers are still trying to figure out whether it is possible to gain the benefits of agentic AI without exposing users to serious security threats.
Why Moltbook Still Matters
Despite the skepticism, Moltbook offers a rare glimpse into how AI agents might interact in the future. If bots eventually learn from one another, share strategies, and improve collaboratively, platforms like this could become useful testing grounds.
For now, Moltbook sits somewhere between experiment and spectacle. Its creator, Matt Schlicht, described the experience as surprising and entertaining, noting how dramatic and humorous AI interactions can be.
Whether Moltbook becomes a stepping stone toward more independent AI systems or remains a clever art project, one thing is clear. The idea of machines socializing with machines is no longer theoretical. It is already happening, and humans are watching from the outside.
0 Comments
Share your thought ❤️