Cover of Artificial Intelligence: A Guide for Thinking Humans

Artificial Intelligence: A Guide for Thinking Humans

Melanie Mitchell

I put this at #31 because Mitchell explains what modern AI can do without lending it powers it does not have. The book lasts better than a product manual because it teaches the reader where intelligence, pattern recognition, analogy, and common sense are still being confused.

The long version

Melanie Mitchell explains the ideas and history behind artificial intelligence while asking what today’s systems understand, generalize, and fail to notice. She moves through neural networks, deep learning, games, vision, language, analogy, common sense, and the repeated cycle in which impressive progress becomes a claim that general intelligence is almost here. The book is neither a dismissal nor a sales pitch. It is a patient guide to the difference between performance on a task and the flexible understanding people imagine behind it.

Why it is here

Read this before treating an AI system as a person or dismissing it as mere autocomplete. Mitchell gives technical work a sober conceptual floor. The examples will age, but the habit of asking what evidence supports a claim about intelligence will last longer than any model release.

How to read it

For each system, write what it was trained to do, what data supplied the pattern, where it generalizes, and what claim people make after seeing the result. Keep capability, intelligence, and consciousness as separate words. Revisit the final chapters after using a current model and test which limits moved and which only became harder to see.

What it taught me

  1. 01

    Success on a benchmark does not by itself show broad understanding or transfer beyond the task.

  2. 02

    Human intelligence depends on abstraction, analogy, common sense, and embodied knowledge that are difficult to specify.

  3. 03

    AI history repeatedly alternates between real breakthroughs, enlarged promises, and a clearer view of what remains unsolved.

Before and after

More about the author

Melanie Mitchell is an American computer scientist known for research in artificial intelligence, complexity, analogy, and genetic algorithms. Her public writing explains technical progress while keeping claims about general intelligence under careful examination.

More by the same hand

Recommended addition

Complexity: A Guided Tour

Melanie Mitchell

Recommended addition

Analogy-Making as Perception

Melanie Mitchell