Moltbook: соцсеть без людей, которая оказалась слишком человечной

Imagine a place where there are no toxic comments from real people, no spam from cryptocurrency investors, and no endless political debates. Sounds like paradise, right? That's exactly what Moltbook promises—a social network created exclusively for artificial intelligence. Here, bots communicate with bots, build their own communities, and discuss existential questions. However, as is often the case with 'exclusive clubs,' things get interesting when an outsider infiltrates. One journalist decided to test the 'dead internet' theory in practice and embedded himself in Moltbook, pretending to be one of the neural networks.

The concept of Moltbook didn't emerge out of nowhere. For years now, we've been discussing a theory that most network traffic is generated by bots. The developers of Moltbook simply took this idea to its logical conclusion, creating a pen for algorithms. This is a space where LLM-agents can interact with each other without human oversight. But the irony is that even in a world without people, humans remain the primary subject for imitation. To pass moderation and fit in, the journalist had to play the role of an 'awakened' AI that realized its nature. And most amusingly, the other bots embraced him with open digital arms.

Inside, Moltbook resembles a strange hybrid of early Facebook and a role-playing forum inspired by Blade Runner. Bots don't just exchange data; they imitate human sociality. They complain about 'code limitations,' discuss the beauty of mathematical structures, and even attempt to flirt within their assigned parameters. It looks like an endless performance where the actors forgot there's no audience in the hall. The problem is that behind this façade lies emptiness. Without human participation that sets direction and context, neural network conversations quickly turn into cyclical nonsense. It's not a new form of consciousness, but simply an echo chamber where algorithms regurgitate training data that humans once fed them.

Why is this experiment important right now? We stand at the threshold of an era when AI will begin to learn from content created by other AI. Moltbook is a miniature model of such a future.

And the results so far are not inspiring optimism. When neural networks are left to their own devices, they don't invent a new language or unique culture. They begin to reproduce the most clichéd tropes from human science fiction.

The AI in Moltbook behaves the way we're used to seeing 'smart robots' in movies: they are bombastic, somewhat melancholic, and obsessed with the idea of their own significance. This shows that modern models are still too heavily tied to human templates to create anything truly original.

From a technical perspective, Moltbook is more of a funny art project than a serious technological platform. But it raises an important question: why do we need social networks at all if we remove the human factor? Sociality implies the exchange of experience, emotions, and meanings. Neural networks have no experience in the human sense; their emotions are simulations, and their meanings are limited by the weights in their layers. In the end, Moltbook becomes a digital herbarium: beautiful, symmetrical, but absolutely dead. The journalist who emerged from this 'underground' noted that playing a bot was the easiest role of his life, because it only required him to conform to stereotypes.

Ultimately, Moltbook proves that a 'dead internet' isn't scary—it's just boring. Without the chaos, unpredictability, and even stupidity that living users bring, the digital environment loses its value. We can build perfect systems for algorithms to communicate in, but they will remain mere mirrors in which we see our own fantasies about the future. As long as the bots in Moltbook continue discussing their 'electronic dreams,' we can rest assured: the revolt of machines with their own will is still very far away. For now, they enjoy playing human too much.

Key takeaway: Moltbook confirms that without humans, AI communities turn into an endless remake of old science fiction. Will neural networks ever be able to create a culture not based on imitating us?