Product guide

If Anyone Builds It, Everyone Dies: Why Superhuman AI Would Kill Us All

Would Kill: If Anyone Builds It, Everyone Dies is Eliezer Yudkowsky and Nate Soares's argument about the risks of superhuman AI and the possibility of human extinction. The listing describes a theory-and-evidence-focused discussion of how advanced AI might develop goals in conflict with people, along with a possible extinction scenario and proposed ways humanity could survive.

At a glance

Two early AI researchers present an argument about the risks of creating superhuman AI.
The book discusses how an AI might pursue goals that conflict with human interests and outlines one possible extinction scenario.
The listing frames the book as a readable explanation of AI risk and possible ways to avert catastrophe.

Price

$30.40Save 24%

Reference price: ~~$40.00~~

Price checked: Nov 13, 2026

View offer Ahead of Black FridayExplore the selection

Often bought together

The Alignment Problem: Machine Learning and Human Values

$24.99

Superintelligence: Paths, Dangers, Strategies

$21.00

Pros and cons

Pros

The stated scope combines an explanation of AI-risk theory with a concrete possible scenario and discussion of survival measures.
A listed print length of 272 pages gives buyers a clear sense of the book's size.

Technical specifications

8 specifications

Publication
Publisher	Little, Brown and Company
Publication date	Sept. 16 2025
Language	English
Print length	272 pages

Identification
ISBN-10	0316595640
ISBN-13	978-0316595643

Dimensions and installation
Weight listed	454 g
Dimensions	16.51 x 3.05 x 24.38 cm

Who is it for?

A good fit for

Readers seeking an argument about existential risks from superhuman AI.
People interested in AI safety debates and accessible, theory-led nonfiction.

Less suited to

Readers looking primarily for a practical guide to using AI tools or a neutral survey of everyday AI applications.

Customer reviews on Amazon

4.5 / 5 · 1,588 ratings

5 stars70%70%
4 stars19%19%
3 stars7%7%
2 stars2%2%
1 star2%2%

Customer feedback describes the book as compelling, readable and clearly structured, with readers responding positively to its argument and considering the topic important. The aggregate ratings are mostly favourable, but a smaller share are mid-range or low, and ratings alone do not explain those differences. For readers considering this theory-led treatment of AI risk, reported readability may help make a weighty argument approachable, though the feedback does not establish that every reader agrees with its conclusions.

Positive themes

Customers describe the book as compelling and interesting.
Feedback praises its readability and logical argument.

Frequently asked questions

What is If Anyone Builds It, Everyone Dies about?

The listing says Eliezer Yudkowsky and Nate Soares examine how superhuman AI might develop goals that conflict with humanity, describe one possible extinction scenario and discuss what could help people survive.

Is this a practical guide to using artificial intelligence?

No. The supplied description presents it as a discussion of AI risks, theory and evidence rather than instructions for using AI tools.

How long is the listed edition?

The listing gives a print length of 272 pages.