What Makes Us Smart: The Computational Logic of Human Cognition

by Samuel Gershman

How a computational framework can account for the successes and failures of human cognition

At the heart of human intelligence rests a fundamental How are we incredibly smart and stupid at the same time? No existing machine can match the power and flexibility of human perception, language, and reasoning. Yet, we routinely commit errors that reveal the failures of our thought processes. What Makes Us Smart makes sense of this paradox by arguing that our cognitive errors are not haphazard. Rather, they are the inevitable consequences of a brain optimized for efficient inference and decision making within the constraints of time, energy, and memory―in other words, data and resource limitations. Framing human intelligence in terms of these constraints, Samuel Gershman shows how a deeper computational logic underpins the “stupid” errors of human cognition.

Embarking on a journey across psychology, neuroscience, computer science, linguistics, and economics, Gershman presents unifying principles that govern human intelligence. First, inductive any system that makes inferences based on limited data must constrain its hypotheses in some way before observing data. Second, approximation any system that makes inferences and decisions with limited resources must make approximations. Applying these principles to a range of computational errors made by humans, Gershman demonstrates that intelligent systems designed to meet these constraints yield characteristically human errors.

Examining how humans make intelligent and maladaptive decisions, What Makes Us Smart delves into the successes and failures of cognition.

Genres: Nonfiction, Science, Academic, Neuroscience, Psychology

224 pages, Hardcover

Published October 19, 2021

Reviews

[Hussain Almohri · Rating 3 out of 5]

The book has a very good collection of concepts about how humans think. Some concepts are trivial to understand albeit the author makes them overly complicated. For example, when explaining psychological experiments with images to learn about cognitive biases, the author fails to first explain the image and then tell the conclusion that could be drawn. In general the language suffers from the academic obscurity in writing, composing confusing sentences. I believe this could have been addressed a lot better from a cognitive scientist.

[Riccardo · Rating 5 out of 5]

A really neat tour through a quite consistent bayesian view of cognition. Highly recommended. With platelet notations, simulations and data re-analysis would make for an amazing course.

[Yudi · Rating 5 out of 5]

This book is a fascinating read and covers so many interesting topics and shows that some unified principles are underlying these many different phenomena.

[RAGHAVENDRA K]

This book offers a computational perspective on perception and decision making. While books like *Thinking, Fast and Slow* discuss bias in human decisions, this book argues that bias is an inevitable outcome of a rational, efficient, machine-like brain. Bias arises because the brain tries to make the best possible inference from limited data and resources. For those in neuroscience, this book provides a Bayesian view of perception and decision making.

I enjoyed this book at several places, and I liked thinking through these problems. At times, I was a bit dissatisfied with the writing style – it included many references and cited results like an academic review paper. To me the second half of the book was the best part. Here is a bit of what stood out:
- **How to never be wrong** – we all strongly believe in some core theories. But when confronted with contradictory evidence, instead of disbelieving the theory, we make additional hypotheses. Why do we do that most of the time? In which cases do we give up on the theories we believe? This chapter discusses these questions within a Bayesian framework.
- **The Frugal Brain** – about how the brain makes an efficient representation of things found in nature.
- **Language design** – about how languages evolved; the core underlying principle is to transfer information with high clarity and low cost.
- **The uses of randomness** – discusses the hypothesis that the brain also does some kind of sampling.

[Anne · Rating 3 out of 5]

Well written for a general audience, with several interesting and important ideas for computational cognitive science thrown in.

[Cole R · Rating 3 out of 5]

I fear he could have made this much easier to understand.