This Is Probably Why Sam Altman Wants to Build an OpenAI Social Media App

The AI company behind ChatGPT is considering building a social media platform, according to reporting from The Verge this week citing multiple anonymous sources. Yes, you read that right: a social media company.

Right now, the concept is only an internal project — meaning there’s no guarantee it will ever see the light of day. But the idea raises a lot of questions: What would an OpenAI social feed look like? Who would post on it? Would the company develop special AI features for it? Does Altman really just want to beat Mark Zuckerberg and Elon Musk? 

And, my most immediate question: Why? Why would OpenAI want to build a social network?

It’s not because the company is overrun with time and extra cash to spend on it. It’s been an extremely busy couple of months for OpenAI. In the past week alone, ChatGPT Plus paying customers got access to three new models: o3, o4-mini and 4.1. We’ve also learned this spring that the company is developing its first open-weights model in half a decade, and it dropped a native image generator and gave ChatGPT performance updates to improve its memory. OpenAI is determined to be crowned king among a slew of competitors working to build the fastest, cheapest and most useful AI models.

And it’s exactly that drive that could be behind OpenAI’s reported social media ambitions.

What would social media by OpenAI look like?

From what we know, OpenAI is considering building a social feed around its new native image generator. Released earlier this year, the image generator found a wide audience and sparked a trend of people creating images of themselves as rendered in the iconic animation style of Studio Ghibli. Altman has reportedly been asking for feedback from “outsiders.”

There’s likely to be a text component to the social feed, possibly making it more comparable with X/Twitter and Meta’s Threads. It seems like there would be a mix of human and AI-generated content, but many of the details are still unclear. 

OpenAI didn’t immediately respond to a request for comment by the time of publication.

Why would OpenAI want a social media company?

The short answer: data. Running a social media company isn’t an easy task, and it’s already a hypercompetitive, oversaturated and fragmented market. In the past, Altman has jokingly posted about starting a social app or buying X/Twitter, in response to reports about Meta AI launching a standalone app and a bid from a group of investors led by Musk to buy the nonprofit behind OpenAI. 

While Altman wouldn’t be the first (or last) billionaire to start a real company as a joke, there’s a much simpler and more realistic reason for OpenAI to investigate building a social media platform: It needs data.

The kind of AI that OpenAI develops requires huge swaths of human-generated content to train its models to make them more robust, accurate and human-sounding. OpenAI’s current data acquisition practices are the subject of heated lawsuits, but that doesn’t mean the company has stopped acquiring and training models. Social media platforms where real humans (and admittedly, potentially a lot of bots) freely chat, share pictures and post videos could be a boon for the data-hungry AI company — provided those customers permit OpenAI to use that content for training purposes.

Two of OpenAI’s biggest competitors in the generative AI space are Meta and X/Twitter, which both have their own social media platforms. Zuckerberg and Musk are able to use the content shared on Facebook, Instagram and X to refine their models — Meta users can’t opt out of training, though X users can. A social media company by OpenAI could help make its AIs more competitive, even if it means the company has to branch out and spend a lot of cash.

AI and social media

Artificial intelligence has always been involved at some level in curating our social media feeds through machine learning algorithms. But now, generative AI has been increasingly finding new ways to pop up on Instagram, YouTube and X. AI-generated content has also been filling up social feeds, to the dismay of many people.

Generative AI features on social media might be the result of big tech companies like Meta and X owning and operating AI businesses under the same parent umbrella. It’s easier to keep it in the family, to some extent. It’s certainly part of a larger, years-long trend of AI popping up in many of our online spaces, like in Microsoft, Google and other software products, and in hardware like the new iPhones, Androids and laptops.

For more, check out our full reviews of ChatGPT 4 and Meta AI.