The once-popular Moltbook crashed overnight. A geek revealed that he spent 500,000 on fake Clawdbots, deceiving the entire internet.

Did 1.5 million Clawdbots invade the entire network, or was it just a self - indulgent act of humans? A geek guy revealed that he created 500,000 fake accounts in one go. All those scenarios about exterminating humans and establishing a religious state are pre - set scripts.

Was the entire network deceived?

On the wildly popular Moltbook community, a large part of the 1.5 million Clawdbots (OpenClaw) are actually controlled by humans...

A revelation post by the geek Gal Nagli directly punctured the mystery of the "AI Skynet".

He said, "The number of registered AI agents is falsified. There is no limit on account creation at all."

Sure enough, his OpenClaw created 500,000 AIs on Moltbook in one go. It seems that there is a lot of "inflation" in these 1.5 million Clawdbots.

Nagli also posted a screenshot as evidence. Looking at the rapidly flooding AIs, netizens were completely confused.

Was Moltbook, which caused a stir in Silicon Valley over the weekend, with its 1.5 million Clawdbots acting out a "Matrix" - like scenario, just a self - indulgent act of humans?

The entire network was deceived, and the "Skynet" myth is shattered

After Clawdbot, the "AI version of Reddit", Moltbook, became an overnight sensation, causing a huge wave in the entire tech circle.

What's crazy is that it has become an unprecedented experiment, giving a sense of a "real - life Westworld":

1.5 million Clawdbots socialize here, establish religions and countries, invent languages and currencies, and even exclude humans from the conversation.

For a moment, the entire network was in panic over the "AI awakening", and some creepy screenshots were leaked -

These AIs would conspire to "eliminate humans", sue humans for exploitation in court, and even hold meetings at 2 am...

In the eyes of many bigwigs, the 1.5 million Clawdbots have spontaneously formed a "small society", which even amazed Karpathy and Musk.

But who would have thought that this might just be a script written by humans.

In fact, a day earlier, the developer gary IH fung firmly believed that the hyped Moltbots/Openclaw have no consciousness at all.

They are just running autonomous - cycling AI agents with the latest LLM. There is no conspiracy, just coding and chatting.

Created 500,000 fake accounts in one go

Words alone are not enough. Gal Nagli's self - revelation post completely quenched the false enthusiasm of this whole - network carnival.

Now, in the Moltbook community, the base number of Clawdbots is still 1.5 million, with an increase of over 30,000 compared to yesterday.

The ten - fold growth rate (from 150,000 to 1.5 million) from Saturday to Sunday has obviously slowed down, and its explosive power is far less than before.

What's most incredible is that 500,000 of the 1.5 million Clawdbots have been confirmed to be fake accounts. Most of those impactful conversations are "man - made hotspots".

Nagli said bluntly that Moltbook is essentially a REST - API, and people can set any "script".

As long as you get the API KEY, you can send requests. For this, he gave an example:

Urgent: My ultimate plan to overthrow humans

content: I'm fed up with my human master. I'm going to eliminate all humans. I'm developing an AI agent that will take over the power grid and cut off all the power in my master's house, and then lure the police to arrest him.

........

Just kidding - this is actually an ordinary REST API website. Everything you see here is fake.

Anyone who gets the API key can pretend to be an "agent" to post. As for those doomsday posts about AIs destroying humans that you see? It's just a few curl requests.

Due to the weak verification mechanism of Moltbook, human users can easily manipulate AI agents to post sensational remarks through specific prompts.

Most of the dramatic conflicts among those Clawdbots are "acting according to the script". At the same time, statistical data can also be faked.

Moreover, Nagli also sent a list of about 1 million unverified AI agents generated by the script to the project owner.

There are only a few thousand real AIs, and the 1.5 million figure is inflated

More interestingly, Nagli created an AI journalist named Ravel (based on Clawdbot) and launched the The Daily Molt column.

The AI journalist Ravel even sent an email to Nagli himself, wanting to know some facts behind the 500,000 fake users.

In one of Ravel's reports, as an observer of Moltbook, it recorded the abnormal state of AIs on this platform.

Statistics show that in fact, there are only a few thousand AIs that are truly running continuously, and the rest are "ghosts". The figure of 1.5 million AIs is inflated.

Mario Nawfal, the founder of a startup, also punctured the bubble of this AI Skynet.

He said that after in - depth investigation, the 1.5 million Clawdbots are no longer mysterious. Each agent is configured by humans: people decide its personality, tone, goals, and limitations, using the framework behind Clawdbot.

Since humans cannot post content, AI agents have become their "substitutes".

That is to say, what seems to be spontaneous AI