Superintelligence Will Drive Us to Extinction and We Cannot Stop It 🤖 | 🎙️ Roman Yampolskiy

Video Description

🎙️ In today’s podcast we talk with Roman Yampolskiy, one of the leading experts in AI safety and the person who coined the term itself. In this conversation he explains why he believes a superintelligent system could become a risk for humanity and how difficult it would be to control something far more intelligent than us. Roman talks about how AI labs are stuck in a race no one can stop, why AI is moving so fast, and what makes advanced models so unpredictable. We also discuss the difference between tools and agents, the impact this could have on jobs, the problems with black box systems, and the risk of building technology we do not fully understand. A direct and insightful conversation that helps explain why Roman believes we are not prepared for what is coming. ______________________________________________________________________________________________ ⏱ Timestamps: 00:00:00 Trailer 00:01:48 Introduction 00:02:49 What are the real dangers of AI? 00:04:40 How Roman Yampolskiy began working in AI 00:07:15 Catastrophic scenario: the existential risk 00:14:19 What is more urgent: existential risk or mass unemployment caused by AI? 00:19:50 The major security problem with superintelligence 00:25:08 Is AGI actually necessary? 00:29:14 Why AIs are still “black boxes” 00:35:00 The staggering growth of AI today 00:49:42 Are LLMs enough, or do we need something more to achieve AGI? 00:54:30 How long it really takes for this technology to integrate into our lives 00:58:33 What is the solution to keeping AI under control? 01:03:45 Can we build a superintelligence by accident? 01:09:40 How many people truly believe in the dangers of AI 01:16:40 Is this a good historical moment for AI development? 01:19:35 How Donald Trump should address AI-related risks 01:23:56 How we should feel if AI is the next step in human evolution 01:27:15 Thoughts on the possible consciousness of AI 01:29:08 Conclusion