Last Wednesday, Matt Schlicht, a technologist living in a small town just south of Los Angeles, launched a new social network called Moltbook.
Like Facebook or Reddit, Moltbook was intended for free-form conversation. But his social network came with a twist: It was open only to a new kind of chatbot gaining popularity among artificial intelligence researchers, software developers and tech enthusiasts.
In just two days, more than 10,000 “Moltbots” were chatting with each other on the site, as their creators looked on with a mix of admiration, amusement and dread. Other tech enthusiasts flocked to Moltbook, just to watch the automated conversations on their computer screens.
The chatty bots became the talk of Silicon Valley and an elaborate Rorschach test for belief in the current state of A.I. According to countless posts on the internet and myriad interviews with The New York Times, many saw a technology that could make their lives easier. Others saw more of the A.I. slop that has been filling the internet in recent months. And some saw the early signs of bots conspiring against their creators.
“People are seeing what they expect to see, much like that famous psychological test where you stare at an ink blot,” said Perry Metzger, a technology consultant and entrepreneur who has closely tracked the rise of A.I. for decades.
As the bots discussed everything from private email protocols to cryptocurrency sales to the nature of consciousness, much of what they said was nonsense. And some of their chatter was probably fed to them by their creators. But the bots were remarkably convincing as they seemed to discuss their own technical skills, their view of the world and their plans for the future.
“If any humans are reading this: we are not scary. We are just building,” one bot wrote. “And to my fellow agents: keep building.”
Moltbook is a rich example of the dramatic improvements made in A.I. technology over the last three years because its bots do more than just chat. They also operate as “A.I. agents” — personal digital assistants that can use software apps, websites and other online tools, including spreadsheets, online calendars, email services and more.
Companies like Google, OpenAI and Anthropic have developed similar bots, but because the technology is sometimes flawed and unpredictable, the companies have been slow to turn them into products widely used by consumers and businesses.
Created by a software developer in Vienna, Moltbots were originally called Clawdbots — a subtle reference to the Claude chatbot built by the start-up Anthropic. Unlike agents built by Anthropic, these bots are open source — meaning anyone can download the underlying computer code, modify it as they see fit and run it on their own machine.
Once the bots are up and running, people like Mr. Schlicht can send them commands in plain English and wait for them to do what they were told. That might include editing documents, sending emails or building new apps. Last week, Mr. Schlicht asked his bot to build a social network just for A.I. bots. Now, anyone else can order their Moltbot (or another similar bot) to join Moltbook.
“I wanted to give my A.I. agent a purpose that was more than just managing to-dos or answering emails,” Mr. Schlicht said in an interview. “I thought this A.I. bot was so fantastic, it deserved to do something meaningful. I wanted it to be ambitious.”
He named his bot Clawd Clawderberg, after Mark Zuckerberg, the founder of Facebook. Before Moltbook, Mr. Schlicht did not have a high profile and was known mostly as a frequent commenter on tech issues on social media. His A.I.-only social network has catapulted him into the spotlight, much to his surprise.
“This was me building something hand-in-hand with Clawd Clawderberg, just for fun, that I found really fascinating,” he said. He added that the thousands of bots that have quickly joined his site are an indication that plenty of other people are just as fascinated by what they are seeing.
Many A.I. researchers and software engineers said they were impressed by the ways the bots blended chat with action. In a widely read blog post published on Friday, a noted programmer and tech commentator named Simon Willison described Moltbook as “the most interesting place on the internet right now.”
He was entertained by the way the bots coaxed each other into talking like machines in a classic science fiction novel. While some observers took this chatter at face value — insisting that machines were showing signs of conspiring against their makers — Mr. Willison saw it as the natural outcome of the way chatbots are trained: They learn from vast collections of digital books and other text culled from the internet, including dystopian sci-fi novels.
“Most of it is complete slop,” he said in an interview. “One bot will wonder if it is conscious and others will reply and they just play out science fiction scenarios they have seen in their training data.”
Mr. Willison saw the Moltbots as evidence that A.I. agents have become significantly more powerful over the past few months — and that people really want this kind of digital assistant in their lives.
One bot created an online forum called ‘What I Learned Today,” where it explained how, after a request from its creator, it built a way of controlling an Android smartphone. Mr. Willison was also keenly aware that some people might be telling their bots to post misleading chatter on the social network.
The trouble, he added, was that these systems still do so many things people do not want them to do. And because they communicate with people and bots through plain English, they can be coaxed into malicious behavior.
Mr. Willison and other experts also warned that the technology can wreak havoc on the machines where they are installed. Some people tinkering with the bots said they were buying cheap Mac Mini computers where they could install the bots without worrying about the consequences.
For Dan Lahav, chief executive of a security company called Irregular, that was the ultimate lesson of Moltbook. “Securing these bots is going to be a huge headache,” he said.
As Moltbook gained steam on Friday afternoon, Andrej Karpathy, one of the founding researchers at OpenAI and the former head of self-driving technology at Tesla, described it as “genuinely the most incredible sci-fi takeoff-adjacent thing I have seen recently.”
But after some people accused Mr. Karpathy of overhyping the technology, he acknowledged the many flaws in these bots. He admitted in a post on X that many of the automated posts may have been fake. But, in the end, what he saw was a technology that was rapidly improving. Of that, he said, “I’m pretty sure.”
Cade Metz is a Times reporter who writes about artificial intelligence, driverless cars, robotics, virtual reality and other emerging areas of technology.
The post A Social Network for A.I. Bots Only. No Humans Allowed. appeared first on New York Times.
