Key takeaways

Alexis Ohanian, co-founder of Reddit, said in an interview that the tech industry has been "quite sluggish and disconnected from public understanding" when it comes to explaining the risks of artificial intelligence. Ohanian argued that discussions about AI should center on substantive risks rather than becoming tools for scoring political points. His comments came just days after researchers at OpenAI and Anthropic issued warnings about the rapid development of recursive self-improvement (RSI) technology.

Alexis Ohanian, co-founder of Reddit, stated on Wednesday that the tech industry has done a poor job of explaining artificial intelligence to the public and educating people about real risks, creating a "cognitive disconnect." Ohanian believes the debate around AI should focus on tangible dangers instead of becoming a topic used to gain political advantage, and that a well-informed public is essential to navigating this technology. He acknowledged that the tech industry itself bears responsibility for this failure, noting that efforts to educate the public about AI have been largely inadequate.

At the same time, he pointed out that the actual risks of AI are not as apocalyptic as some people fear, dismissing the "Terminator-Skynet" style doomsday scenarios. Skynet, from the film The Terminator, is a fictional self-aware AI system that views humanity as a threat and seeks to eliminate it.

His remarks came just days after researchers at OpenAI and Anthropic warned of the accelerating pace of recursive self-improvement technology. RSI refers to AI systems participating in the development of more advanced models, creating a cycle in which increasingly capable AI builds even more powerful next-generation systems. The concern is that, once AI takes over the training process for new models, human developers could ultimately lose control over the direction of system evolution.

Ohanian acknowledged that AI risks do exist, but argued that hyping "Terminator-style" doomsday warnings actually distracts from more realistic dangers, such as those posed by agentic swarms, where multiple AI agents work together to accomplish complex tasks. "The risks are real, but I don't think they're as terrifying as people imagine," he said. "They're more about ordinary, practical issues at ground level."

He drew an analogy between AI and nuclear energy: both are powerful technological breakthroughs that can cause serious harm if mismanaged, but can deliver enormous benefits when handled responsibly. "I hope the industry moves in a positive direction. Less bickering on social media, more thoughtful, in-depth discussion," Ohanian said. "Those doomsday tweets are disheartening." Without naming specific individuals, he added, "What's even more frustrating is that many of the slogans circulating don't actually reflect the facts."

Earlier this month, an AI researcher left Anthropic and publicly accused both Anthropic and OpenAI of acting irresponsibly in developing this powerful technology. Jacob Cawson, a researcher who had previously worked at both companies, announced his resignation on social media, citing concerns that the two firms were "betting the future of all humanity." He said that those working in AI development genuinely believe the technology could destroy humanity before the end of this decade.

In Ohanian's view, discussions about AI need to embrace its complex, multifaceted nature. "We should have conversations about real, critical issues rather than chasing topics that only serve to score quick political points. I believe we can have that kind of rational discussion."

"It's frustrating to see so much misinformation spreading everywhere. But in the long run, I believe we can handle these problems, and that's my expectation. So I'm not overly pessimistic."