No. 175 of 500 No. 35 in Mind, Brain & Philosophy

What Is Intelligence?

Lessons from AI About Evolution, Computing, and Minds

By Blaise Agüera y Arcas

Print length
624 pages
Publisher
The MIT Press
Reading level
Intermediate
ISBN-13
978-0262049955
Published
September 23, 2025
Written for
Scientifically minded readers interested in what large language models reveal about minds, life and evolution.

Synopsis

Blaise Agüera y Arcas, a Google vice president who founded the company's Paradigms of Intelligence research team, argues that the success of large language models should not have come as a shock. If brains evolved to predict what happens next, then a network trained to predict the next word could plausibly develop something like general intelligence. He pushes the claim further: life itself is computational, replication and evolution are forms of computing, and intelligence is defined by what a system does rather than the material that composes it. The book moves from the origins of life and symbiogenesis through cybernetics, learning, other minds, transformers and consciousness, and closes by asking whether AI marks a new evolutionary transition.

Agüera y Arcas supports the argument with experiments from his own team, including simulations in which short self-copying programs arise and evolve, and with literature from neuroscience, physics and machine learning. He takes positions many researchers dispute, including that some current AI systems have a credible claim to intelligence and even consciousness. Reviewing it in the Los Angeles Review of Books, Patrick House praised the book's patience with definitions while noting its hurry to showcase computational tools. Readers gain a coherent functionalist framework for thinking about minds, machines and evolution, along with a clear statement of one side of the debate over whether AI systems understand.

Scorecard

Editorial score

How it scored

Editorial score: 70.3 / 1005 weighted criteria

  • Influence30%

    19.5/30
  • Authority25%

    21.3/25
  • Practical value20%

    10.0/20
  • Currency15%

    13.5/15
  • Readability10%

    6.0/10
  • Editorial score

    70.3/100

How the rank was computed

Rank No. 175 of 500 comes from a composite of 57.8: observable signals 26.1 of 55 points plus editorial review 31.7 of 45 points.

All 11 inputs
InputPoints
Observable signals26.1/55
Reach (Amazon ratings volume)7.6/16.5
Consensus across best-of lists8.3/13.8
Awards and bestseller evidence0.0/8.3
Reader rating, sample-adjusted5.8/8.3
Amazon sales rank1.6/5.5
Recency2.8/2.8
Editorial review31.7/45
Author authority9.6/11.3
Influence8.8/13.5
Practical value4.5/9.0
Readability2.7/4.5
Currency6.1/6.8

Each criterion is scored from 1 to 10 and weighted, and the points add up to the editorial score. The rank combines observable signals (55 percent) with this score (45 percent). Read the methodology.

Verdict

An ambitious, wide-ranging argument that prediction and computation unite life and intelligence, rewarding patient readers with original research and bold claims.

Strengths

  • Original research from a leading AI lab
  • Synthesizes biology, physics, neuroscience and computing
  • Clear stance in the AI-understanding debate

Keep in mind

At more than 600 pages and with sweeping claims about consciousness, it demands patience; readers skeptical of functionalism will find many conclusions contestable.

Key takeaways

  1. Prediction may be the core function of brains and even of life, which helps explain why next-word prediction produced unexpectedly general AI systems.
  2. Intelligence is functional and multiply realizable: it depends on what a system does, so it can arise in carbon or in silicon.
  3. Life is computational; replication, variation and evolution can be modeled as code, and complex systems grow through the merger of simpler ones.
  4. Minds become necessary when modeling other minds matters, linking social life, theory of mind and the evolution of intelligence.

Notable facts

About the author

Blaise Agüera y Arcas

Blaise Agüera y Arcas is a vice president and fellow at Google and its CTO of Technology & Society, where he founded the Paradigms of Intelligence team. At Google he led work on on-device machine learning and the invention of federated learning. He earlier founded Seadragon Software, became a Microsoft Distinguished Engineer, and wrote the books Ubi Sunt and Who Are We Now?

Questions

What is What Is Intelligence? about?
Blaise Agüera y Arcas argues that prediction is the core of intelligence and of life itself, and that intelligence depends on function rather than material. He traces this idea from the origins of life through evolution, brains and social cognition to large language models, arguing that today's AI continues the same processes that produced biological minds.
Is What Is Intelligence? worth reading in 2026?
Yes, for readers who want a serious argument about what large language models reveal about minds. Published in 2025, it engages directly with current AI and won a 2026 PROSE Award category prize. Its length, abstraction and contested conclusions about AI consciousness make it a demanding read, and skeptics of functionalism will want to read its critics alongside it.
Do you need a technical background to read What Is Intelligence?
No formal background is required, because the author introduces computation, biology and neural networks as he goes. Still, the book is long and moves across physics, evolutionary biology, neuroscience and machine learning, so readers comfortable with demanding popular science will find it easier. The Antikythera program also hosts a companion website organized by chapter.

Sources

As an Amazon Associate, Best AI Books earns from qualifying purchases. Best AI Books is sponsored by Iternal.ai, whose founder and CEO, John Byron Hanby IV, wrote The AI Strategy Blueprint, our #1 ranked book. How we rank books.