← Home
The Diary of a CEO with Steven Bartlett · May 22, 2026 · 36m

Most Replayed Moment: AI Safety Expert Predicts The Next 20 Years! Will It Really Take All Jobs?

Roman Yampolskiy
YouTubeSpotifyApple PodcastsShow Notes

Dr. Roman Yampolskiy, who coined the term 'AI safety' in 2010, discusses the trajectory of artificial intelligence over the next two decades. The conversation centers on which jobs AI and robotics are most likely to replace, whether superintelligence is inevitable, and whether humanity has lost control of AI development. Yampolskiy explores the structural challenges of regaining safety oversight and what work itself might look like in an age of autonomous systems.

This summary was generated from show notes and public descriptions, not from a full transcript review. Details may contain inaccuracies.

Curious

•
Superintelligence as Loss-of-Control Event
Once AI systems reach superintelligence—exceeding human cognitive capability across all domains—the window for human control and oversight may have permanently closed.
ClaudeChatGPTGemini

Highlights

•
Driverless Cars as First Wave of Job Displacement
Transportation and driving jobs represent the most immediately vulnerable labor category to AI displacement, with driverless cars emerging as near-term technology.
ClaudeChatGPTGemini
•
Humanoid Robots as General-Purpose Labor Replacement
Humanoid robots designed to perform human tasks across multiple domains represent a potential inflection point where job displacement accelerates beyond specific sectors.
ClaudeChatGPTGemini

Editorial

•
The Structural Problem of AI Safety Governance
Current AI development operates under competitive, profit-driven incentives that work directly against safety-first practices, making centralized safety governance nearly impossible without major economic restructuring.
ClaudeChatGPTGemini
•
Work as Identity and Purpose May Disappear
If AI and robotics eliminate most human labor, the social structures built around work—identity, purpose, status, daily structure—face simultaneous collapse without a coherent replacement system.
ClaudeChatGPTGemini

Misc

✧Yampolskiy has published over 100 papers on AI dangers and existential risk
✧First person to formally coin 'AI safety' as a field in 2010
✧Focus on driverless cars and humanoid robots as near-term displacement vectors
Was this useful?