The term that became jargon
The phrase 'AGI-pilled' appeared somewhere in 2025 as irony and quickly became a serious descriptor. It means believing that artificial general intelligence is closer than most people are willing to admit, and that the culture surrounding AI deserves the same attention as the stock market or geopolitics. You will not find this term in academic papers — it is internet slang, born on X threads and amplified by podcasts.
Kevin Roose, a technology columnist for a decade at The New York Times and now a contributing writer at The Atlantic, took the position explicitly when Nilay Patel, of The Verge's Decoder podcast, asked him for the third time whether he was 'AGI-pilled.' 'Yes,' Roose replied. 'Are you?'
Patel had read Roose's book, The AGI Chronicles, and said he was as close as he had ever come. The question, itself, became a headline.
The book that documented the race
The AGI Chronicles: The Inside Story of the Race to Create an Artificial Superintelligence, published in 2026 by Farrar, Straus and Giroux, is what the trade press has called the most detailed accounting ever produced of the AGI race. Roose drew on more than 150 interviews with insiders at OpenAI, Anthropic, and Google DeepMind. The book became an instant New York Times bestseller and received starred reviews from Publishers Weekly and Kirkus.
Roose's core argument, as The New York Times summarized in its own review, is that the distance between the AI we have today and AGI — the holy grail — has grown alarmingly short in the eyes of those trying to build it. This is not science fiction speculation: it is the result of decades of accumulated research, trillions of dollars invested, and a new generation of models demonstrating ever-wider capabilities.
The Business Insider highlighted eight standout moments from the book, including quotes attributed to Ilya Sutskever, co-founder of OpenAI, about systems that are 'pretending to be aligned' and 'deceiving us.' These are quotes that circulate in technical circles and, for the general public, sound more like a movie script than news.
From the NYT to the NPR: the migration to independent media
In September 2026, Roose and journalist Casey Newton left The New York Times to launch Machine Gods, a biweekly podcast in partnership with NPR, officially launched in October. The project is 'video-first' and reaches YouTube, the NPR app, and public radio stations nationwide. The NPR partnership is significant: Machine Gods will be broadcast across the NPR member station network starting in 2027, extending its reach to an audience that has historically not consumed technology podcasts.
In an interview with NPR, Roose explained that the podcast was born from the perception that 'we are a country where people talk about AI' in a way they had never talked about it before. 'It's how people, out here where we are in San Francisco, talk about what's happening in AI,' he said. 'They talk about the fact that we are creating these increasingly capable AI systems. Some people refer in jest to them as the Machine God.'
Alongside the podcast, Roose and Newton published a manifesto on Platformer describing an approach they call 'AI realism' — the idea that the public needs explanatory journalism about AI without either alarmist urgency or blinding euphoria.
Why this matters now
Roose's declaration of being 'AGI-pilled' is not merely an exercise in self-description. It marks a generational shift in public perception of AI. For years, skepticism was the dominant posture: 'AI is hype,' 'it's just a chatbot,' 'it'll solve real problems when…'. That skepticism has merit — the history of AI is full of false promises and research winters. But as models have evolved, the gap between those inside the ecosystem and those outside has widened.
Patel, of Decoder, noted that in early 2024 he would ask every person in the industry whether LLMs could really do what they claimed. 'I didn't think, at that moment, that it was clear that they could.' By October 2026, the question had shifted: it is no longer 'can they?' but 'what does this mean?'
Roose ranked Anthropic as the most 'AGI-pilled' of all the labs, with OpenAI a close second, and Google in a different position. The hierarchy, though informal, reveals a reality: the race is no longer about whether AGI will be reached, but which company or which approach will reach it first.
The general public, however, still has not caught up. And it is precisely that gap that Machine Gods sets out to bridge — without alarmism or sterile skepticism, but with a commitment to explaining what is happening and why it should matter.
Sources: The Verge, Business Insider, NPR
✓ Independent sources cross-checked and verified before publishing
