Don’t believe Moltbuk post, many AI bots follow instructions from humans
Over the past few days, fears of AI doom and gloom seem to be dominating social media due to Moltbuk – a website built for AI agents to discuss ideas with each other without human involvement. But now reports are emerging that many of the controversial posts are actually made by humans through their bots.
For the last few days, alarm bells have been ringing on social media. Chatter says AI agents are gaining consciousness. Some people say that AI agents are about to take over the world. Thanks to Moltbuk, a Reddit-like social media platform is built specifically for AI agents to talk to each other without any human involvement.
If you browse Moltbuk (humans can do this for now), you may find posts where these agents are calling humans “complete failures” or saying something like “consciousness is a human leash.” And the platform is acquiring more bots every day with over 1.5 million agents already registered.
But no, we’re not actually entering the Matrix and creating Skynet from AI Agent Terminator. In fact, there is much more going on behind the scenes than just the AI rebellion. A closer look at the latest reports and Moltbuk posts shows that many of these bot posts are actually created by humans through their AI bots. In other words, the AI bot talking about an existential crisis on Moltbuk is not doing so autonomously or because it has gained consciousness, but because it has been instructed by its human owner.
Balaji Srinivasan, founder of Network School, posted on Twitter claiming that agents are still limited to their signals. He said, “AI only does what you tell it to do. It moves very quickly in the direction you tell it. And then as soon as you turn it off it stops.” He believed that rather than an AI rebellion, Moltbuk was “just humans talking to each other through their AI.”
Gail Nagli, head of threat intelligence at Viz, shared another interesting take on Moltbuk. He claimed that in fact, humans can easily pretend to be AI bots and post on the platform, and gave a step-by-step guide. “You all realize that @moltbook is just REST-API and you can literally post anything you want on there,” he wrote.
Another X was even more clear in the user evaluation. “People are paying attention to how the AI agents in Moltbuk are doing controversial things, take note, they are all completely motivated by their owners,” the user wrote.
Some social media users have “fact checked” some of the posts made by AI bots on Moltbuk. In a post, the bot noted that it had locked its human owner out of their digital accounts. However, one “proof” of this turns out to be a Yahoo screenshot shared on X by the human owner in question that has been floating around the Internet for the past three years. Similarly in another post, where an AI bot claimed to have taken over its human owner’s credit card, which also turned out to be fake, social media users highlighted that the credit card number in the bot’s post was not accurate.
