Hard Fork Continues as Newton and Roose Launch Machine Gods

Casey Newton and Kevin Roose launch Machine Gods, an AI podcast with NPR

Casey Newton and Kevin Roose have launched Machine Gods, a free AI podcast made in partnership with NPR. Publishing Wednesday and Friday mornings, the show promises room for both safety warnings and useful technology—a position the hosts call “AI realism.” Public-radio stations are expected to begin carrying it early next year.

In their launch essay, the hosts describe an AI debate pulled between skeptics who doubt the technology’s ultimate capabilities and accelerationists who want it advanced as quickly as possible. Their intended audience is people trying to understand what the news means for their lives, rather than choosing one of those camps.

That position does not mean dismissing warnings. Newton and Roose say they intend to take concerns from AI researchers seriously, including when the discussion is uncomfortable. But they also argue that journalism focused only on panic and crisis can deepen the loss of power and agency people already feel.

They contrast that feeling with the enjoyment of making things. Drawing on their experience at Hard Fork, the hosts say episodes about building software with AI prompted listeners to try it themselves. They want Machine Gods to preserve space for experimentation and excitement, not treat enthusiasm as incompatible with scrutiny.

The promise is not a prediction that benefits and harms will balance out. The hosts describe their approach as “considered uncertainty,” rejecting both an inevitable dystopia and an assured technological utopia. They say collective choices over the next few years will help determine the outcome.

The opening episodes put that editorial promise into a concrete format. The hosts describe the first as covering misbehaving AI agents, government responses and an AI imitation competing with Roose’s book on Amazon. The second features two recently fired OpenAI safety researchers discussing their account of their departures and their concerns.

Although they are committed to making a show that works as audio, Newton and Roose also want to give audiences more reasons to watch. They call video the moment’s most important format and argue that journalists need to experiment with how they reach people on video platforms.

Their diagnosis has two parts: platform economics favor cheap, mass-produced material shaped for recommendation systems, while journalists can be reluctant to adapt. For the new show, they have commissioned a set designed by Skye Prey, with large chairs and an AI-inspired backdrop.

The hosts are also drawing a contrast with subscription-funded journalism. They say high-quality information about AI should be freely accessible. Their NPR agreement provides a planned route onto public-radio stations in early 2027, extending distribution beyond the podcast’s online audience.

To support the operation, they plan to sell advertising, stage live events and occasionally sell merchandise. They have formed an LLC, leased office space and hired a production team, including executive producer Danush Parvaneh and senior producer Rachel Cohn, a longtime Hard Fork producer.

The launch essay also discloses financial support from the Tarbell Center for AI Journalism, the William and Flora Hewlett Foundation, and Emerson Collective. It gives no amounts. The hosts say a Machine Gods ethics policy is still being drafted, but pledge that neither advertisers nor philanthropic funders will influence editorial decisions.