Reviving the Blue Network

Apr 21, 2025

∙ Paid

Something changed recently. There’s been talk that OpenAI is working on a social network like X. That’s big news for the following reasons:

Due to ChatGPT’s current popularity, it could quickly grow to rival X.
Due to network political tribalization, a built-in user group (the blue network) is ready and willing to adopt it as an alternative to X, driving daily usage.
This would be the first social network built on AI, as a platform rather than an incorporated feature.

Let’s dig in.

Popularity

Let’s compare the popularity of ChatGPT and X.

Momentum. Open AI is the most downloaded app in the world, far surpassing X; 45 million (ChatGPT) vs. approximately 20 million (X) every month.
User base. ChatGPT has 800 million monthly active users, a reasonable estimate of its installed base, compared to X’s approximately 600 million.
Activity. X users are far more active than ChatGPT users, for obvious reasons. ChatGPT has 122 million active daily users, compared to X’s 436 million, a significant gap.

Building a social network on an AI platform would make the network unique for the following reasons.

It would incorporate AI functionality into all posts. Grammar and spell check out of the box. Analysis of the quality of the post (i.e., it could estimate the popularity of the post and what could be done to improve it). Dynamic posts that update themselves, seamless image creation or retrieval based on the post, talking head video shorts (your avatar), etc.
Built in ‘safety’ and moderation at a level never seen before (it’s already doing that for ChatGPT users). Twitter and Facebook employed thousands of people to moderate their networks. OpenAI’s network would be capable of moderating every post in real-time, not just based on keywords, but also through the contextual analysis of each post, without requiring a large staff like Twitter or Facebook.
AI-generated feeds and dynamic user experiences. A social network built on an AI would allow people to construct AI-generated feeds that automatically populate and interact with users on any subject. It would also enable users to tailor their feeds in real time, focusing on a specific group of people or a particular topic. Lastly, it would also be capable of creating a corpus of ‘verified’ information (aka ‘authorized by experts’) that it would use to discredit or verify posts (instead of the dynamic approach X uses).

The arrival of OpenAI’s social network is well-positioned to leverage network political tribalization. Let’s put this into context.