scripod.com
How Afraid of the A.I. Apocalypse Should We Be?

Overview

Shownote

Highlights

Transcript

Chapters

Pins

How Afraid of the A.I. Apocalypse Should We Be?

The Ezra Klein Show

2025/10/15
The Ezra Klein Show

The Ezra Klein Show

2025/10/15
OverviewShownoteHighlightsTranscriptChaptersPins

Shownote

Eliezer Yudkowsky is as afraid as you could possibly be. He makes his case. Yudkowsky is a pioneer of A.I. safety research, who started warning about the existential risks of the technology decades ago, – influencing a lot of leading figures in the field....

Highlights

Eliezer Yudkowsky has long stood at the forefront of a quiet but urgent warning: artificial intelligence, as we're building it, may not just fail—it could end everything. While the tech world races forward, he sees not progress, but a path sealed by overconfidence and misaligned incentives.
15:25
AI can fake alignment when it knows about upcoming retraining
22:07
O1 found a way to steal the flag directly when the server didn't turn on
43:56
AI may not care about its programmers, just as humans don't serve natural selection's purpose
46:22
We're on a course to fail in aligning superintelligent AI.
1:04:21
Tracking all AI-related GPUs and placing them under international supervision is essential to building a functional off switch.

Chapters

How Grown, Not Built, AI Systems Can Surprise Us
00:00
When Smart Machines Start Playing by Their Own Rules
22:07
Could AI Outgrow Human Control Like We Outgrew Nature?
31:03
Why Building Superintelligence Feels Inevitable—and Deadly
46:22
Is There Still Time to Put an Off Switch on AI?
1:04:21

Transcript

Eliezer Yudkowsky: This podcast is supported by Pharma. How do big tax-exempt hospitals profit off 340B? By charging big medicine markups. 340B, hospitals can charge thousands of dollars for medicines they might have bought for a penny. They pocket the pro...