Sciencepulse

Why the People Building AI Are Among Its Loudest Worriers

AI·November 7, 2026

AI is settling into daily life faster than many people can track, and the race to build more capable systems is accelerating rather than slowing down. In a recent episode of Scientific American's podcast Science Quickly, tech journalist Kevin Roose looks at where that race stands and why it unsettles some of the people running it.

The conversation centers on artificial general intelligence, or AGI, a loosely defined goal of building systems that can match or exceed human ability across a wide range of tasks instead of one narrow job. Roose's reporting points to an unusual pattern: a noticeable share of the people working on these systems are also among the most outspoken about the risks. Their concern is not a side debate inside the industry. For them, it is part of the picture of what they are building.

The episode also turns to the wider stakes, including what rapid progress could mean for society, for safety, and for the long-term future of humanity. Those questions remain open, and the discussion treats them that way rather than offering tidy answers. For most people, though, the stakes already feel concrete. AI tools now show up in search, workplace software and phones, so choices made by a small number of companies and researchers shape products that millions use every day.

Roose explores the topic at greater length in his new book, The AGI Chronicles: The Inside Story of the Race to Create an Artificial Superintelligence, published by Farrar, Straus and Giroux in 2026. For readers who want the fuller account of how the people involved think about the technology and its dangers, the book is the natural next step after the podcast.