El Podcast
E167: Nuclear Rockets, AI Agents & Science Hype | RealClear Science’s Ross Pomeroy
Episode Summary
Steven Ross Pomeroy, science writer and editor of RealClearScience, joins the show to talk NASA’s lost nuclear rocket program, the promise of space mining, and why we still haven’t gone “all in” on advanced propulsion. He also dives into hidden AI adoption at work, “scienceploitation” and health hype, information overload in the smartphone era, and what neuroscience really says about speed reading and how we learn.
Episode Notes
Steven Ross Pomeroy, Chief Editor of RealClearScience, joins the podcast to discuss NASA’s abandoned nuclear propulsion programs, the future of AI and white-collar work, the rise of “scienceploitation,” and how information overload is reshaping human cognition.
GUEST BIO:
Steven Ross Pomeroy is a science writer and Chief Editor of RealClearScience. He writes frequently for Big Think, covering space exploration, neuroscience, AI, and science communication.
TOPICS DISCUSSED:
- NASA’s nuclear propulsion program (1960s–1970s)
- Why nuclear rockets were abandoned
- Differences between chemical, nuclear thermal, and nuclear electric propulsion
- Using the Moon as a launch hub
- Moon-landing skepticism & conspiracy thinking
- The future of space mining
- AI adoption trends & hidden usage
- Agentic AI vs chatbots
- Job displacement: white-collar vulnerability
- Higher ed, skills, and career advice
- “Scienceploitation” and how marketing hijacks scientific language
- Immune-system myths & quantum woo
- Information overload and Google/AI-driven forgetting
- Critical thinking in the AI era
- The myth of speed reading
- How vocabulary and deep engagement improve comprehension
MAIN POINTS:
- NASA had functional nuclear-rocket tech in the 1960s, but political priorities, budget cuts, and waning public interest ended the program.
- Nuclear thermal rockets are ~2x as efficient as chemical rockets; nuclear electric propulsion could unlock deep-space exploration and mining.
- Space mining is technologically plausible, but its economic impact (like crashing gold prices) creates new problems.
- AI adoption is much higher than official numbers—many workers use it quietly and off the books.
- Companies see low ROI today because they’re using simple chatbots, not advanced “agentic” systems that can take multi-step actions.
- White-collar jobs — not blue collar — are being automated first.
- Scienceploitation hijacks scientific buzzwords (“quantum,” “immune-boosting,” “natural”) to sell products with no evidence.
- We process 74 GB of information per day, roughly a lifetime’s worth for a well-educated person 500 years ago.
- Speed reading works only by sacrificing retention; the real way to read faster is to build vocabulary and deep attention.
- Skepticism, not cynicism, is the core skill we need in the AI-mediated media environment.
TOP 3 QUOTES:
- “It would’ve been harder to fake the moon landing than to actually land on the moon.”
- “Companies aren’t getting ROI from AI because they’re only using chatbots. The real returns come from agentic AI — and that wave is just beginning.”
- “We now process 74 gigabytes of information a day. Five hundred years ago, that was a lifetime’s worth for a highly educated person.”