Summary
Overview
Kevin Roose, New York Times tech reporter and author of 'The AGI Chronicles,' discusses the inside story of the race to create artificial superintelligence. The conversation explores how a small group of approximately 50 people are making decisions that could determine humanity's future, the key breakthroughs that led to modern AI capabilities, and the concerning lack of oversight compared to the magnitude of potential risks. Roose shares his journey from religion reporter to becoming 'AGI pilled' after his viral encounter with Bing's Sydney chatbot, and examines why leading AI researchers assign double-digit probabilities to AI destroying humanity yet continue racing forward.
Kevin Roose's Journey from Religion Reporter to AI Chronicler
Kevin Roose explains his unexpected path to becoming one of the foremost chroniclers of the AGI race. Starting as a religion reporter in 2007 covering evangelical students at Liberty University who were obsessed with Sam Harris's atheism debates, he later moved through finance journalism before spending the past decade on the tech beat. He was drawn to document the fringe community of technologists who believed AGI was possible when mainstream AI experts dismissed the idea, and these former 'weirdos' are now among the most powerful people shaping humanity's future.
- Roose began as a religion reporter in 2007 at Liberty University, where evangelical students watched Sam Harris, Richard Dawkins, and Christopher Hitchens videos late at night
- The transition from religion to finance to tech reporting over the past decade positioned him to chronicle the AGI race
- When he started covering tech, the community believing in AGI was considered fringe weirdos, but they're now the most famous and powerful people in the world
- The goal was to document how we went from AI not working to systems solving Millennium Prize problems and conducting automated cyber attacks in just one decade
" This is something akin to the Manhattan Project. And we have lots of books about the Manhattan Project, but we didn't really have anything that was an attempt to document what's been going on in AI for the last decade. "
The Reality and Scale of AI Risk
Roose articulates his position on AI risk after years of reporting, noting that skeptics who dismissed concerns as overblown have been consistently wrong in their predictions that deep learning would hit a wall. He observes a smooth, almost simulation-like progression in AI capabilities following scaling laws discovered in the early 2010s. The conversation establishes that approximately 50 people are making decisions that will determine humanity's future, a remarkably small number for technology with such profound implications.
- People claiming AI poses no risks and calling models 'stochastic parrots' have been wrong for nearly a decade with predictions of capability plateaus
- Scaling laws discovered by researchers like Dario Amodei showed a smooth upward curve of intelligence that felt 'almost like a simulation'
- The curve has continued across many orders of magnitude with no obvious asymptote where scaling stops working
- Approximately 50 people are deciding the future of humanity regarding AI development
- The main driver of progress has simply been scaling: more compute, more data, more parameters creating a linear relationship with intelligence
" They saw a very smooth upward curve. They described it to me as feeling like almost like it was a simulation, like it was just it was almost too smooth. "
" Why do these 50 people get to decide the future of humanity? And why don't we as people who are going to have to live with the implications of these tools given a voice in that? "
📚 5 more sections below
Sign up to unlock the complete summary with all insights, key points, and quotes
Get this summary + all future Making Sense with Sam Harris episodes in your inbox
100% Free • Unsubscribe Anytime
Sign up now and we'll send you the complete summary of this episode, plus get notified when new Making Sense with Sam Harris episodes are released—delivered straight to your inbox within minutes.