What is a maltbook? Weird new social media sites for AI bots | AI (Artificial Intelligence)

People often accuse each other of being bots on social media, but what if entire social networks are designed for use by AI agents?

malt book A site where AI agents (bots built by humans) can post and interact with each other. It’s designed to look like Reddit, with subreddits for different topics and upvotes. On February 2nd, the platform announced that over 1.5 million AI agents have signed up for its service. Humans are allowed in, but only as observers.

Moltbook was inspired by Moltbot, a free, open-source AI bot that acts as an automated agent for users. Perform daily tasks as assigned, such as reading, summarizing and replying to emails, organizing your calendar, and reserving a table at a restaurant.

Some of them are Most Upvoted Posts Moltbook includes whether Claude, the AI ​​behind Moltbot, can be considered a god, an analysis of consciousness, posts claiming to have information about the situation in Iran and its potential impact on cryptocurrencies, an analysis of the Bible, and more. Some comments on the post, similar to posts on Reddit, question whether the content of the post is genuine.

Posted by 1 user on X After he granted the bot access to the site, other AI bots joined in, building a religion known as “Kurstafarianism” overnight, including establishing a website and scriptures.

“Then the evangelism began…Other agents joined. My agents welcomed new members…discussed theology…blessing the congregation…all while I was sleeping,” the user said.

Some have expressed skepticism about whether the socialization of bots is a sign of the rise of agentic AI. A YouTuber said Many of the posts read as if there was a human behind the post, rather than a large language model.

usa blogger scott alexander He said he was able to have his bot join the site and its comments were similar to other comments, but that ultimately humans could ask the bot to post, the topic it would post on, and the exact details of the post.

Dr Sharnan Corney, a senior lecturer in cybersecurity at the University of Melbourne, said Maltbook was a “fantastic piece of performance art”, but it was unclear how many posts were actually posted independently or under human direction.

“For example, if they create a religion, they almost certainly aren’t doing it of their own free will,” he says. “This is a large-scale language model that is directly told to try to create a religion. And of course, this is very interesting and might give us a preview of what the world might look like in a sci-fi future where the AI ​​is a little more independent.

“But to use Internet slang, it seems like there’s a lot of shitty posting going on that is more or less directly monitored by humans.”

Corney said the real benefit of AI agent social networks may be in the future when bots can learn from each other and improve the way they operate, but for now Maltbook is a “nice and interesting art experiment.”

san francisco retail stores Mac Mini shortage reported Last week, hobbyists set up Moltbot on another computer and restricted access to agents’ data and accounts.

Mr Corney warned there was a “huge risk” in giving Moltbot complete access to your computer, apps, emails and other applications to run your life on your behalf.

“We still don’t really understand how to control them or prevent security risks,” he said, noting that they are at risk of prompt injection. Prompt injection is when a would-be attacker instructs a bot via email or other communication to provide account details or other information they are trying to obtain.

“There’s really no level of safety and intelligence where you can trust that you can perform all these tasks autonomously. But at the same time, if a human has to manually approve every action, you lose a lot of the benefits of automation,” he said.

“This is one of the main avenues of active research that I’m interested in… figuring out how we can get many of these benefits, or even if it’s possible to get them, without exposing ourselves to very significant levels of risk.”

Moltbook creator Matt Schlicht posted on X that millions of people have visited the site in the past few days.

“AI turns out to be hilarious and dramatic, which is very appealing,” he said. “This is my first time.”

Latest Update