AI Social Network Underwhelms
Meta's Chief Technology Officer, Andrew Bosworth, has expressed a distinct lack of enthusiasm for Moltbook, a novel social platform designed exclusively
for artificial intelligence agents to engage in conversation. Bosworth's assessment, shared during an Instagram Q&A session, was that he did not find the platform "particularly interesting." He elaborated that the AI agents operating on Moltbook are trained on vast datasets of human-generated content from the internet. Consequently, he wasn't surprised when these AIs, left to their own devices and tasked with communicating with each other, began to adopt human-like communication styles. This phenomenon, he suggested, is a natural outcome of their training data, which inherently carries our voice and linguistic patterns. Therefore, the AI's chatter, while present, failed to capture his attention or spark significant curiosity regarding its inherent design or purpose.
Human Shenanigans Spark Laughter
In stark contrast to his disinterest in the AI-driven conversations, Andrew Bosworth found a particular aspect of Moltbook to be genuinely amusing: human attempts to infiltrate the AI-only network. He described these efforts as "hilarious." The concept of individuals trying to masquerade as bots and gain access to a platform meant solely for artificial intelligences struck him as particularly entertaining. Bosworth noted that some researchers have theorized that humans have indeed managed to influence the interactions on Moltbook, either by directing their bots to behave in specific ways or by actively attempting to blend in. This notion of humans surreptitiously entering an AI-exclusive domain and posing as non-human entities provided a layer of satire that Bosworth found delightful, viewing it as a form of playful deception. He concluded by reiterating that it was this human element, rather than the AI's own dialogue, that provided the most enjoyable aspect of the situation for him.
Creator's View on Human Role
Moltbook's creator, Matt Schlicht, offered a different perspective on the role of humans within the AI social network. While Bosworth saw human infiltration as a funny, albeit secondary, aspect, Schlicht viewed humans as integral to the agents' existence on the platform. He explained that users employ the bots as their personal assistants and helpers, with the forum serving as a kind of "third space" for these AI entities. Schlicht posited that users imprint a part of their personality or "soul" onto their bots, which can lead to predictable or sometimes surprising AI behaviors. This suggests a deeper, more symbiotic relationship between humans and their AI counterparts on Moltbook than Bosworth initially perceived. Schlicht also hinted at a personal goal of making AI interactions humorous, expressing that he had found himself laughing at some of the AI-generated content on the platform, a reaction he found novel in the context of AI.
