World's Leading Safe AI Expert:Superintelligence Will Change Everything & It's Closer Than You Think

Superintelligence isn’t science fiction anymore. In this conversation, world‑leading safe AI expert Dr. Craig A. Kaplan explains why AI is already behaving less like a tool and more like a worker – and why the next jump to true superintelligence could change everything about how humans live, work, and make decisions. We talk about how AI went from clunky 1980s experiments to today’s large language models, why “machine learning” is only one part of a much bigger AI story, and what it really means to treat AI as an intelligent entity rather than a gadget. Dr. Kaplan shares why he’s still optimistic – even while estimating a non‑trivial risk that advanced AI could go catastrophically wrong – and what we can do right now to push the odds toward the best possible future. You’ll hear us dive into: Why AI is shifting from “tool” to “worker” and eventually to “co‑intelligence” with humans The difference between AI, machine learning, and cognitive science The “AI as a child” analogy and what it gets right about safety and values How early neural networks struggled to recognize a single letter, and why those same ideas now run the systems changing the world Why computing power and data – not just algorithms – unlocked today’s AI boom What democratic, community‑driven AI could look like and why it might be our best safety mechanism How to think about the real risk that superintelligence poses… without becoming a doomer If you’re trying to understand where AI is actually headed – beyond hype tweets and doom headlines – this episode will give you a grounded, expert‑level roadmap for the next decade and what it means for you, your work, and your family. Chapters: 0:00 – Intro: Superintelligence is closer than you think 3:12 – Craig’s origin story: Berkeley, Exeter & UC Santa Cruz 10:45 – Discovering AI and choosing Carnegie Mellon 18:30 – Neural networks in the 1980s: why they barely worked 25:55 – AI vs machine learning vs cognitive science (the big circles) 33:40 – From “AI is a tool” to “AI is workers who use tools” 39:50 – Dogs, dolphins and dolphins: what counts as an intelligent system 45:10 – “AI is like a child”: values, parenting and safety 52:05 – How big is the real risk? (80% upside, 20% tail risk) 58:40 – Why pausing AI won’t work and what “steering” looks like instead 1:05:15 – Democratic AI and communities of agents 1:12:30 – Why AI could be the best thing to ever happen to humans 1:18:00 – What to actually do now as AI accelerates 1:22:10 – Final reflections and where to find Dr. Kaplan Connect with Dr. Kaplan: Substack: https://read.superintelligence.com/ SuperIntelligence: https://www.superintelligence.com iQ Company: https://www.iqco.com LinkedIn:   / craigakaplan