The Stories We Are Told About AI: Introducing the “AI: Tools or Gods?” Podcast

Every week, a new claim about artificial intelligence enters the public conversation. AI will take your job. AI will save democracy. AI is already smarter than you. Some of these claims are true. Some are marketing. Some are mythology dressed as inevitability. Most are doing political work that nobody has bothered to name.

The work of information labs has always been to name that work. That is what the research does. That is what the book does. And starting now, that is what the podcast does.

AI: Tools or Gods? takes the dominant narratives about artificial intelligence and puts them in the room with someone who knows better. It launches at https://informationlabs.transistor.fm/shows and wherever you listen to podcasts.

What the podcast does

Each episode opens with a provocation — a claim, a framing, or an assumption that is currently shaping how people think about AI. Then a guest who has spent years inside the problem takes it apart.

The format is the same every time. One expert. One uncomfortable argument. A lively conversation that goes somewhere the usual AI coverage does not.

Every episode closes with the same two questions: What is the one thing decision-makers most consistently misunderstand about AI right now? And if you could change one thing about how the public conversation about AI is conducted, what would it be? The answers, accumulated across guests and episodes, are the signal.

The first four episodes

The opening episodes make the stakes concrete.

Maria Sukhareva, principal AI expert at a DAX company and author of the AI Realist newsletter, opens the series on the gap between AI hype and AI reality. Her diagnosis is more uncomfortable than most. That gap is not a communication problem. It is structural, driven by incentives that better technical literacy will not fix. She breaks down the last-mile problem of AI progress, explains why the majority of organisations are seeing no return on their AI investment, and makes the case that the agentic AI narrative is, in her words, a next-token predictor dressed in a marketing campaign.

Pia Lauritzen, Danish philosopher and author of Questions (Johns Hopkins University Press), examines what we lose when AI becomes the answer machine and nobody asks who controls the questions. Her research shows that almost 80 percent of the questions people ask are what and how questions, and only around 6 percent are why questions. AI systems are structurally built to accelerate the first two. The concern is not just that they cannot answer why. It is that by providing instant, frictionless answers to everything else, they reduce the discomfort that normally prompts us to ask why at all.

Payal Arora, Professor of Inclusive AI Cultures at Utrecht University and author of From Pessimism to Promise (MIT Press), takes on the AI doom narrative directly: it is a Western export. Her fieldwork across India, Brazil, South Africa, China, and the Middle East shows that outside the places where AI policy is written, the dominant mood toward technology is not fear but hope. Pessimism is a privilege for those who can afford to live in despair. The rest of the world has different questions about what AI might do for them, and those questions are largely absent from the governance rooms.

B Cavello, Director of Emerging Technologies at Aspen Digital and co-organiser of the Public AI Network, examines something most AI policy coverage ignores entirely: that the legal definitions of AI are not neutral descriptions of a technology. They are choices about power. Every definition embeds a decision about who is covered, who is accountable, and where enforcement authority lies. Small differences in language, sometimes only a few words, can substantially alter who bears legal responsibility and who escapes it. Definitions are not a technical problem. They are a political one, dressed in the language of precision.

Why this, and why now

The podcast is an extension of AI Tools, Not Gods (BTF Press, 2026, foreword by Brewster Kahle), published open access at aitoolsnotgods.org. The book argues that mythological framing of AI embeds false assumptions into regulation, and organises its case around four myths: Magic, Madness, Heaven, and Sin. The podcast applies the same discipline in real time, episode by episode, as new narratives emerge and new guests test them.

The argument is not that AI is unimportant. It is that the stories told about AI are consequential, and they are mostly untested. Unchallenged myths become policy. Policy shapes what gets built, what gets regulated, and who bears the cost.

Where to find it

New episodes are available athttps://informationlabs.transistor.fm/shows or Youtube and on Apple Podcasts, Spotify, and all major directories. Subscribe there, or sign up for the newsletter to be notified when new episodes drop.

The first episode will come out next week, so make sure to subscribe in order not to miss it.