The tech industry has been “tone deaf” in explaining AI and properly informing the public about real risks, Reddit co-founder Alexis Ohanian told CNBC on Wednesday.
Ohanian said the debate around AI should focus on substantive risks rather than issues that can be used to score political points, arguing that an informed public is essential to navigating the technology. He acknowledged that the tech industry has contributed to the problem by doing a “pretty tone deaf job” of explaining AI, while suggesting that its risks are more mundane than the “Terminator and Skynet” scenarios some fear.
Skynet is a fictional self-aware AI system in the “Terminator” movies that saw humans as a threat and sought to wipe them out.
The comments come days after warnings from researchers at AI labs OpenAI and Anthropic of the rapid advancement of the technology known as recursive self-improvement, or RSI.
RSI refers to AI systems helping develop more advanced models, potentially creating a cycle in which increasingly capable systems contribute to building even more capable ones.
The concern is that if AI systems take over the process of training new models, their human developers could eventually lose control over how those systems evolve.
Ohanian argued that while AI risks are real, apocalyptic “Terminator”-style warnings can distract from real concerns such as agent swarms, when multiple AI agents work together to perform a complicated task.
“There are risks, I do think they’re more benign, or they’re more banal than people are thinking,” he said.
He compared AI to nuclear energy because both technologies are powerful scientific advances that can cause serious harm but also deliver major benefits if managed responsibly.
“I’m hopeful this goes in a healthy direction, and it’s just going to require fewer tweets and more, like, thoughtful conversation,” Ohanian said.
“The doomer tweet was discouraging,” he said without specifying, adding it was “frustrating” when “the slogans are actually not real things.”
Earlier this month, an AI researcher quit his job at Anthropic and accused the company and OpenAI of acting irresponsibly when it comes to developing the powerful technology.
Jacob Coxon, who has worked as a researcher at both companies, said on social media that he resigned because he was concerned that Anthropic and OpenAI are “gambling with our lives.” He said the people building AI “earnestly believe that it could kill us all by the end of the decade.”
For Ohanian, nuance is crucial. “We should have a discussion about the important, like real stuff, instead of these things that maybe score some quick political points… I have faith that we can have this discussion.”
“It frustrates me because I see all this misinformation flying around, but I do think in the long run, we [will] get it right, and that’s what I’m hoping happens. So I’m not terribly worried about that part of it.”