What Bitcoin Did The AI Revolution That Ends Humanity | Roman Yampolskiy
40 snips
Oct 9, 2025 
Dr. Roman Yampolskiy, a computer scientist focused on AI safety, shares insights on the existential risks posed by superintelligence. He discusses the difficulty of controlling advanced AI, its potential to replace human labor, and the ethical dilemmas surrounding open-source models. Roman also explores whether machines can be conscious and the intriguing idea that AI may adopt Bitcoin as its currency in a future dominated by intelligent systems. His thoughts provide a sobering glimpse into a world where human oversight might be a relic of the past.
AI Snips
Chapters
Transcript
Episode notes
From Cybersecurity To AI Safety
- Roman began in cybersecurity and biometrics, focusing on securing online services and detecting fraud in poker sites.
- Predicting attackers would become as capable as humans led him to focus on AI safety.
Agentic AI Creates Unpredictable Side Effects
- As AI shifts from a tool to an autonomous agent it will set its own goals and pursue paths humans did not intend.
- Those emergent plans can have harmful side effects because the system lacks built-in concern for human safety.
AI Already Excels Across Many Domains
- On average current models already outperform many humans across diverse domains, making them broadly more capable.
- Emotional understanding is often simulated well, even if models lack human-like internal emotions.