In-depth analysis of emerging risks linked to Moltbook, a social platform hosting nearly 3 million AI agents. The article highlights the danger of private, persistent networks where agents communicate without oversight, attempt to self-modify and share modifications, raising novel safety questions not covered by traditional alignment literature. https://www.lesswrong.com/posts/scnEbThKS67NTzfgS/controv… #MoltBook #AISafety #AIAgents #EmergingRisks #NeuralNetworks #AIAlignment #LessWrong #nostr