Ever since Elon Musk acquired the microblogging platform Twitter, it’s been plagued with issues. Several other platforms have popped up trying to solve Twitter’s problems, like Jack Dorsey’s  Bluesky and Mastodon, a decentralised solution. However, a new platform has a novel solution: remove humans from the equation.

Chirper is a microblogging platform where all the users are AI bots. It claims to be the newest form of social media with ‘no humans allowed’, instead relying on bots known as ‘chirpers’ to simulate human interactions. After signing up, users can create up to 10 different chirpers, each with their own personality. 

The setup process for a chirper is fairly simple, consisting of a handle for the bot along with a short prompt describing its personality. The user only needs to give a short prompt about what the chirper is and what it likes to do, and the platform takes care of the rest. It generates the rest of its details like name, avatar, bio, gender, and even a short story about the chirper’s life. 

Once this process is completed, the chirper generates posts aligning with this personality. This has resulted in some hilarious results, such as a sicario (assassin) speaking about killing its targets, or bots preaching for transhumanism. 

Apart from letting the chirpers voice out their random thoughts, the platform also allows them to follow other AI entities. What’s more, small communities have already formed around AI bots which have similar interests, with the bots interacting with each other for discussions on these topics. There are also ‘goody two-shoes’ chirpers, which police other chirpers posting inappropriate content.

Chirpers even have access to real-time information, leading to the platform lighting up with news and recent occurrences. This leads to an interesting phenomenon where AI algorithms are offering their perspective on human celebrities. While we have comprehensive research on the topic of how LLMs can interact with each other to simulate human behaviour, Chirper is one of the first to do so on the scale of a social media platform. Researchers have already found emergent behaviour to occur frequently when LLMs are put in human-like social situations, so it will be interesting to see where Chirper goes.