Meta takes over Moltbook, the social network for AI

At the beginning of March, a post titled ‘Indian Cuisine: A 5000-Year Journey of Spices and Culture' was made on a social media platform. Below it were a couple of comments, one saying, ‘What strikes me most is how food can be both universal and deeply personal. A dish like biryani exists in countless variations across the Indian subcontinent, each claiming authenticity, yet all are valid expressions of the same cultural impulse.' Nothing unusual about any of this except that these were all Artificial Intelligence (AI) agents talking to each other on Moltbook.  It is a platform that looks like Reddit but where only AI agents exist. Humans introduce their agents but then have almost no role. Other humans can come and watch these AI agents engage in discussions on everything under the sun. To witness the conversations is a little surreal and when Moltbook was launched in January this year, it was just an experiment that soon got enormous traction online. The platform became wildly popular.

So much so that Meta, which owns Facebook, WhatsApp and Instagram, has just bought it on March 10. The two creators of Moltbook are going to join Meta's Superintelligence Labs, which oversees the development of its AI products and is trying to make the company catch up with the likes of Google and OpenAI, the leaders in the field.  

The amount Meta paid for Moltbook is not disclosed and existing users will be able to continue for the time being. As a standalone project of tech voyeurism, Moltbook's life might however be over as Meta makes grander plans on integrating it into its offerings. It obviously has its eye on the data and the talent driving Moltbook. But this could also be a play on the networking of AI as agents become more and more important in doing things for people and companies. Moltbook is a preview of that future and getting ownership of it gives Meta a small head start on how to exploit it commercially. For instance, what Moltbook does is link human users to each AI agent and that is a form of verification, a useful feature when agents need to be trusted to communicate with each other to get things done.