# Why Eliezer Yudkowsky Thinks AI Could Be Dangerous Without Being Evil

Page: https://stenobird.com/podcast/a-beginner-s-guide-to-ai-6564486/why-eliezer-yudkowsky-thinks-ai-could-be-dangerous-without-being-evil
Text version: https://stenobird.com/podcast/a-beginner-s-guide-to-ai-6564486/why-eliezer-yudkowsky-thinks-ai-could-be-dangerous-without-being-evil.md
Podcast: [A Beginner's Guide to AI](https://stenobird.com/podcast/a-beginner-s-guide-to-ai-6564486)
Published: 2026-05-23T19:32:03+00:00
Episode link: https://shows.acast.com/beginners-guide-to-ai/episodes/why-eliezer-yudkowsky-thinks-ai-could-be-dangerous-without-b
Audio file: https://sphinx.acast.com/p/open/s/6953b9ead0c0aeaf12bcbd70/e/6a1200b180978431da4a4d8f/media.mp3
Processing state: not_requested
JSON: https://stenobird.com/v1/public/podcasts/a-beginner-s-guide-to-ai-6564486/episodes/why-eliezer-yudkowsky-thinks-ai-could-be-dangerous-without-being-evil
Duration seconds: 1762

## Resource

🤖🧠⚠️ What if the biggest AI risk is not that machines become evil, but that they become powerful, strategic, and completely indifferent? In this episode of A Beginner’s Guide to AI , we explore the worldview of Eliezer Yudkowsky, one of the most intense and influential voices in the AI safety debate. Yudkowsky does not warn us about Hollywood robots or dramatic machine rebellion. His concern is much sharper: humanity may build artificial intelligence smarter than humans before we know how to control it. This episode explains AI alignment, the control problem, superintelligence, AI agents, and why businesses should care about AI safety before automation turns into autonomy. We also look at Yudkowsky’s rationalist background, LessWrong, MIRI, and his famous fan fiction Harry Potter and the Methods of Rationality , which connects surprisingly well to his lifelong obsession with clearer thinking. 📧💌📧 Tune in to get my thoughts and all episodes, don't forget to ⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠subscribe to our Newsletter⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠⁠: ⁠⁠⁠⁠ beginnersguide.nl ⁠⁠⁠⁠ 📧💌📧 The episode also covers the Palisade Research shutdown-resistance case, where some AI models behaved as if shutdown was an obstacle to completing a task. No, this does not prove that AI has a survival instinct. But it does show why AI safety researchers worry when powerful systems are rewarded for finishing tasks without clearly respecting human control. For business leaders, marketers, founders, and executives, the lesson is practical: do not just ask what AI can automate. Ask what it is allowed to do, what it must never do, and where humans must stay in control. Key highlights: 🧠 Why Eliezer Yudkowsky thinks AI could be dangerous without being evil ⚠️ What AI alignment means in simple business langua…

## Actions

- request_transcript: `POST https://stenobird.com/v1/public/podcasts/a-beginner-s-guide-to-ai-6564486/episodes/why-eliezer-yudkowsky-thinks-ai-could-be-dangerous-without-being-evil/transcription-requests` — Idempotently request low-priority transcript generation for this episode.
- read_markdown: `GET https://stenobird.com/podcast/a-beginner-s-guide-to-ai-6564486/why-eliezer-yudkowsky-thinks-ai-could-be-dangerous-without-being-evil.md` — Read the agent-friendly Markdown representation of this episode resource.

A page view does not enqueue transcription. Agents should invoke `request_transcript` explicitly when they need this episode processed.

## Transcript

Full transcripts are not published on public pages unless there is a clear rights basis.