Episode Details

Send us Fan Mail

A robot vacuum that locks you out of your own house sounds like a joke until you realize it’s a perfect model of modern AI risk: a system with a goal, the ability to act, and zero built-in concern for collateral damage. We take apart the Hollywood story of “evil AI” and focus on the more realistic danger, amoral optimization with agency. Along the way, we draw on Stella Pop’s argument about what happens when every business has access to the same AI, and why the real finish line isn’t “smarter than humans.”

From there we move into the biggest leverage point AI already has: the abstractions that run civilization. Money, laws, corporate structures, credit scores, and compliance rules are informational systems, and AI is uniquely equipped to navigate them at scale. That’s where “the map is not the territory” becomes more than a slogan. AI can master dashboards and proxies, optimize KPIs, and still miss the clanking engine that a grounded human can hear. We connect that to predictive processing, phantom limbs, and the key difference between embodied intelligence and models trained on representations.

Finally, we outline a practical path forward for AI in business and the future of work: augmented judgment. Let the machine generate, search, and simulate, while we bring taste, context, empathy, and moral imagination to decide what should happen and own the consequences. If this reframes how you think about artificial intelligence, share it with a friend, subscribe, and leave a review with the most important human skill you think we can’t afford to lose.