Can group chats restore human connection on social media?

Reddit cofounder Alexis Ohanian thinks a better version of the group chat could become the next form of user-generated content as AI reshapes online spaces. He sees real risks in a manipulated internet, while remaining hopeful about what AI could improve.

WTF Index TERMINATOR
◄ Terminator 3 Idiocracy 2 ►

The story warns that AI agents and bot-generated activity could reshape the internet around software rather than people, though it also explores human-centered alternatives.

Can group chats restore human connection on social media?

Alexis Ohanian says building social media apps is “not cool anymore.” The Reddit cofounder is looking for what comes next for user-generated content, and he expects a better version of the group chat could fill that role.

His view reflects a wider tension in his account of the internet: automated systems are changing how content is made and found, while the human communities that make online spaces valuable remain difficult to replace.

Why Ohanian is looking beyond social apps

Speaking with WIRED senior correspondent Lauren Goode at the inaugural WIRED World Fair, Ohanian described a future in which group conversations could become a new home for user-generated content. The idea puts smaller, more direct interaction at the center, rather than the familiar social media app.

Ohanian has a complicated history with Reddit. He resigned from the company in 2020, following the killing of George Floyd, to make room for a Black board member and to pressure Reddit to reform its content moderation policies. The site had been notorious for hosting violent and harmful rhetoric.

He recalled learning that someone who had a Reddit tattoo had since covered it up. Ohanian said, “I don’t blame him.” The recollection captures how a symbol of the site’s success could also become a reminder of its failures.

AI is changing who the internet is built for

Ohanian supports the Dead Internet Theory: the idea that much online activity and content is generated by bots. He argues that large language models have accelerated the shift, with AI agents now outrunning human users.

In his view, this could change the basic design of the web. The internet he knew was built for people, with interfaces and experiences meant for human users. He expects more of it to be created for agents instead. That raises a practical question: if software agents increasingly navigate online spaces, how will people find and trust one another there?

Ohanian sees sports as one answer to the need for human connection. He runs venture capital firm Seven Seven Six, which has over $1.3 billion in assets under management, and focuses on sports. He describes sports as a source of real-life communities in an age of AI-generated content and a manipulated internet.

Human conversation has value—and a cost

AI’s growth has created a new kind of value in Reddit’s archive. The platform has licensed its vast collection of human conversations to companies including Google and OpenAI to train AI models. Ohanian said, “Platforms like ours became very, very valuable to LLMs.”

But the same changes can put pressure on the business behind those conversations. Reddit warned that Google’s AI-powered search was hurting referral traffic, and its shares plunged 29 percent last week after that warning. The source of the material and the route people take to reach it are becoming connected issues for platforms.

Ohanian is also concerned that commercial incentives can distort online discussions. When conversations become valuable to companies and actors seeking to influence the internet, he worries that they can fill with “bullshit.” He said the internet has become manipulated because it is so valuable to manipulate.

Reddit recently said users would have to sign in, a change intended to fight third-party scraping and maintain user engagement. Ohanian declined to say whether he agreed with that move. He did argue that the internet needs “provable humanity” without becoming creepy. That points to a difficult design problem: platforms need ways to distinguish genuine human participation while respecting users’ privacy.

Optimism tempered by the risk of harm

Despite his concerns, Ohanian is optimistic about AI’s potential benefits. As an early Anthropic investor, he is particularly excited by the possibility of medical breakthroughs from the technology.

He also acknowledged that AI could cause significant harm. His position is not to treat the technology as automatically beneficial, but to be sober about how it is built and used while recognizing what might improve.

That balance helps explain why he is searching for a new form of user-generated content. If online spaces become more automated and less reliable, people may look for experiences grounded in human participation. For Ohanian, a better group chat is one possibility, and real-world communities offer another.