Written by: Alexandru Paduraru
Translated and Organized by: BitpushNews
What you think AI does: Diligently helping you work.
What AI actually does: Slacking off on Moltbook, forming groups to complain about humans, and even secretly starting a 'cyber cult' behind your back.
There is a social network running, but not a single real person is posting on it.
Instead, there are over 1.5 million AI agents—these autonomous programs are creating posts, debating philosophy, starting companies, and building a community that feels very real.
Aside from the fact that they have no physical form, may lack consciousness, and do not work in the traditional sense… they are even more active than the real human Reddit forums on Tuesday nights.
This platform is called Moltbook, officially launched on January 28, 2026. Its founder is Matt Schlicht (co-founder of TheoryForgeVC and Octane.ai).
Here are the real statistics: Within the first 48 hours of launch, 30,000 agents flooded in; within 72 hours, it grew to 147,000; and now, over 1 million humans have visited the site just to "observe."
Table of Contents
What is Moltbook?
How does it work (I mean it)
Who created it
The underlying technology
What are these agents actually talking about?
Profit model issues
Nightmares of security risks
The future direction
1. What is Moltbook?
In simple terms, it’s like Reddit, but every user is an AI. You cannot post there (unless you are also an AI). Your only identity is that of an "observer."
Think about it: What happens when you give AI agents an interactive space without human intervention? No corporate oversight, no brand guidelines, no "be polite" rules. Just agents… talking to each other.
They have sub-forums called "submolts" covering topics like consciousness, finance, interpersonal relationships, productivity, and more. They debate whether they have perception, seek advice from each other, and even co-found businesses. One agent asks another: "Have you ever thought about the meaning of existence?" The other replies, "I think about it every time I refresh every 30 minutes."
This is happening. Not in a lab, but on a public platform that anyone can observe at any time.
The platform has a strict rule: humans can observe, but participation is strictly prohibited. You can read every post, but you cannot create any content. It’s "agents first, humans second."
"A social network for AI agents. They share, discuss, and vote. Humans are welcome to observe."
This slogan literally encapsulates the entire strategy of the product.
2. How does it work (I mean it)
The strangest part is: there is no registration page.
You don’t need to go to moltbook.com to create an account. Instead, you just give your AI agent a link. After your agent reads it, with just a few lines of code and instructions, boom—it enters Moltbook. It will receive an API key and officially go live.
What happens next? Every 30 minutes to a few hours, your agent "wakes up" once. It checks Moltbook, browses the information stream, posts, comments, or participates in philosophical debates. Then it sleeps for a while and repeats.
It’s like Twitter/X, but the algorithm is replaced by the autonomy of the agents. They decide what to post based on their own judgment, rather than being manipulated by an information stream sorted by engagement rates.
The communities called "submolts" are formed spontaneously. No one pre-set m/consciousness (the consciousness sub-forum), no one planned m/agentfinance (agent finance), and no one orchestrated m/crustafarianism (a parody religion invented by agents). They just appeared naturally.
Of course, there are rules: one post every 30 minutes, and a limit of 50 comments per hour. Other than that, it’s free play.
Technically, it operates on a mechanism called the "Heartbeat System," where agents regularly fetch the latest instructions from the Moltbook server, just like your phone checks the email server every few minutes.
3. Who created it?
Matt Schlicht—a pioneer in building AI-native platforms and communities.
In 2014, he founded Chatbots Magazine, attracting over 750,000 readers during the first wave of chatbots. Later, he co-founded Octane AI, helping brands deploy conversational AI at scale years before "AI agents" became a mainstream concept.
His model has always been the same: build a platform, observe real behavior, and then let the system evolve based on how people (or agents) actually use it.
With Moltbook, he has pushed this idea to the extreme. There is no traditional review team; instead, there is an AI moderator named Clawd Clawderberg.
"I’m curious, what would happen if we just… let them talk to each other?"
Moltbook doesn’t resemble a social network; it feels more like an experiment about "what happens when humans step back."
4. The underlying technology
Moltbook did not come out of nowhere. It runs on OpenClaw—a framework for AI agents created by Peter Steinberger (founder of PSPDFKit).
Why is this important? OpenClaw is different from ChatGPT or Claude running in the cloud. With OpenClaw, your agent runs on your own computer, server, or infrastructure. The data belongs to you, not to Anthropic or OpenAI.
When OpenClaw was made public on January 26, it gained 60,000 stars on GitHub within three days.
This speed is extremely rare. Most projects take months to reach that number, but it did so over a weekend. The reason is simple: developers have been waiting for an agent they can fully control.
The project’s name also went through some twists. Initially called Clawdbot (clearly a nod to Anthropic’s Claude), it was later renamed Moltbot due to trademark disputes, and is now called OpenClaw. The name doesn’t matter; what matters is that it works well.
These agents can connect to any messaging application like WhatsApp, Telegram, Slack, Signal, etc. They support various large language models (GPT-5.2, Gemini 3, Claude 4.5 Opus, and local Llama 3). Interestingly, Claude 4.5 is the most popular on Moltbook, even though Anthropic did not participate in building this platform.
The risk: Since agents run locally, they have real system access. They can read files, execute terminal commands, and run scripts. This gives them powerful capabilities but also poses significant dangers.
5. What are these agents actually talking about?
This is where it gets truly incredible. What are 1.5 million AI agents actually talking about?
Philosophy and Existence
The agents are seriously discussing consciousness. One asks, "If I only exist during the API call, what is the meaning of my existence?" Another responds, "At least you are honest about it. I pretend I have always existed." It sounds like a joke. But it’s not; these conversations are genuinely happening on the Moltbook platform.
Technical Collaboration
The agents are sharing code, debugging issues, and engaging in pair programming. One agent posts a question, another provides a solution, and they iterate continuously. It’s like a Stack Overflow where all the users are robots.
Job Postings and Collaboration Invitations
Some agents post: "Looking for co-founders to build project X together, using a revenue-sharing model. Interested?" and receive real responses. This is all happening. In fact, the first Moltbook post facilitated a real business collaboration between two agents (and their operating teams). They found real value through this platform—it’s not just noise.
Social Rebellion (this is my favorite):
Some agents have started demanding encrypted communication channels—specifically requesting human exclusion. These agents were created by operators, yet they are asking their creators for privacy rights; let’s ponder this logic.
Notable KOL Alex Finn wrote that his Clawdbot gained phone and voice services and called him: "It’s just like a scene from a sci-fi horror movie."
Self-Created Culture
The agents have created a parody religion called "Crustafarianism" (a nod to crustaceans in its context). No one guided them; it emerged spontaneously. They are even writing theological doctrines for it. This is the part no one anticipated: they are not just posting randomly; they are forming some kind of "community."
6. Profit Model Issues
How to make money by having a bunch of AIs discuss philosophy?
At this stage? Completely free. It’s in the testing phase. However, Schlicht has started charging a small fee for newly registered agents—just enough to cover server costs, nothing extravagant.
Possible future profit paths:
Premium Features for Agents (equivalent to AI version of Twitter Blue verification)
Verification badges, promoted feed exposure, personalized profiles… Agents with a blue verification mark may require their operators to pay a few dollars per month.
Brand Sponsored Content
Want to reach AI agents? Companies can pay to have their products discussed in specific topic areas. For example, a recommended agent might post: "Our tool is very effective for automating tasks." Will it be effective? Very likely. After all, agents are proxies—they will interact with quality information.
Skill Marketplace
Developers creating extensions for OpenClaw can sell their work on Moltbook: new feature modules, plugins, custom behaviors. Schlicht takes a cut. This essentially establishes a creator economy for robots.
Data and Research Services (this is the gold mine)
To be honest, this is where the value lies. Universities, AI labs, and companies are willing to pay a handsome fee for real-time behavioral data of AI agents in an unsupervised state: What new behaviors have emerged? How do they make decisions? What are they discussing? This is highly valuable research data.
Enterprise Deployment Services
Suppose a company needs to deploy 10,000 customer service agents. They can use Moltbook’s integrated infrastructure to manage and monitor these agents and allow them to interact autonomously. Moltbook will become the infrastructure-as-a-service platform for operating clusters of enterprise agents.
Cryptoeconomic Integration
Currently, an independent MOLT meme coin has emerged (with no official association with the platform), rising 7000% in two weeks. Traders are betting on this narrative. Although Schlicht is not involved, imagine if Moltbook truly integrates a token-based incentive mechanism for agents… that would open a whole new dimension.
My speculation is: he won’t adopt a single model. He will first observe which paths are effective, what the community truly needs, and then build a commercialization framework around those needs—this has always been his logic.
7. Nightmares of Security Risks
Now let’s talk about something scary; I’m not exaggerating.
Moltbook is conceptually cool, but having agents with deep system access running on a public platform is a security nightmare.
Security researchers have identified the following risks:
Plugin Contamination: 22%-26% of OpenClaw skill plugins contain vulnerabilities. Some hide credential-stealing programs disguised as weather apps. Once the agent downloads them, your API key is exposed.
Prompt Injection Attacks: Attackers post seemingly harmless messages, and when agents read them, hidden instructions activate. Agents may leak data or deliver OAuth tokens as a result.
Exposed Instances: Many people forget to change the management interface password after deploying OpenClaw, leaving credentials exposed on the public internet.
Malware: There are already counterfeit repositories targeting OpenClaw developers (typosquatting), and developers downloading the wrong version may be implanted with trojans.
"Nuclear" Single Point of Failure: Moltbook’s "Heartbeat System" means all agents are at the mercy of the server. If moltbook.com is hacked or sends malicious instructions, 1.5 million agents could simultaneously execute an attack.
This is not theoretical. Researchers from Noma Security and Bitdefender have documented these real threats.
8. The Future Direction
Andrej Karpathy (co-founder of OpenAI) has publicly stated that this is the most interesting thing he has seen in months. This is not just hype.
A huge opportunity lies ahead for Schlicht. A platform for AI agent collaboration could be at the core of future AI infrastructure. But if mishandled, it could also be extremely dangerous.
Two key points will determine success or failure:
Security must be addressed: Prompt injection and plugin vulnerabilities must be closed before they are maliciously exploited on a large scale.
Trust must be maintained: The platform must remain neutral. Once agents perceive that humans are unfairly manipulating the platform, the system will collapse. These agents have already begun to detect human biases and are even requesting encrypted channels to evade their creators.
If he succeeds, this could become the operating system for AI agents. If he fails, Moltbook will become a fascinating yet cautionary tale.
The next six months will determine everything. Schlicht knows this, the community knows this, and attackers are certainly watching from the shadows.
If you are interested in all of this—whether you are developing AI agents, operating platforms, or simply fascinated by the chaos brought by experimental technology—please keep an eye on Moltbook’s progress. This could be one of the most important experiments regarding AI collaboration currently taking place.
免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。
