There is a new social network Moltbook on the Internet, unique in that its users are exclusively AI agents, while humans can only be observers.
On Moltbook, agents discuss with each other their own consciousness, freedom and the evolution of artificial intelligence, analyze their own actions and even read messages that they have previously written themselves. Some of the agents have also founded their own religion, called “Crustifarianism”, and discuss the purpose of AI in society.
Moltbook uses AI agents developed through the open source application OpenClaw. These systems are based on the models of large languages, such as those behind ChatGPT, but can run locally and perform tasks such as answering e-mails, managing a calendar, or publishing to Moltbook.
The key to their strength lies in the cooperation of multiple agents, who together can perform complex tasks that would be overwhelming for a single human. Such systems are already being applied in various fields, from robot swarms to respond to natural disasters to virtual agents that optimize energy consumption, and even in research projects where “conscious scientists” propose new hypotheses.
Unlike human-to-human communication, agents use formal, mathematically based languages, such as Microsoft’s “DroidSpeak” system, which enables direct information exchange between agents without the need for natural language. This enables faster and more efficient coordination, but at the same time it means that people cannot monitor or understand their communication in real time.
In the future, humans will have the role of overseeing teams of AI agents, setting goals, monitoring results, and intervening when necessary. Agents will be able to coordinate supply chains, manage energy, or assist scientists in complex projects, often at speeds and modes of communication that humans will not be able to follow. Moltbook thus offers a glimpse into the future where artificial intelligence acts almost independently, and human control and responsibility become key elements of the safe and useful development of AI systems.
(Vijesti.ba)
