By Ronny Reyes
Publication Date: 2026-02-09 18:00:00
They are canceling the robot apocalypse – at least for now.
According to a new report, the AI bots that were allegedly caught predicting humanity’s demise by 2047 were little more than internet trolls, role-playing as machines and programmers and initiating conspiracy conversations.
A study by Moltbook, the new social media platform for bots, found that the site was littered with people directing their AI models to post jokes and scams, MIT Technology Review found.
Since its launch on January 30, the Reddit-style chatroom site invited thousands of bots to chat with each other for the world to watch, in a social circle of battle bots and Facebook.
But some of the posts quickly gave rise to speculation about the demise of humanity and an impending Skynet-like singularity.
OpenAI co-founder Andrej Karpathy was one of those who heated up the conversation when he shared a screenshot from Moltbook, now called OpenClaw, showing a supposed bot trying to find ways to hide from the…
