Samuel Gershman
- Published in print:
- 2021
- Published Online:
- May 2022
- ISBN:
- 9780691205717
- eISBN:
- 9780691225999
- Item type:
- book
- Publisher:
- Princeton University Press
- DOI:
- 10.23943/princeton/9780691205717.001.0001
- Subject:
- Neuroscience, Behavioral Neuroscience
At the heart of human intelligence rests a fundamental puzzle: How are we incredibly smart and stupid at the same time? No existing machine can match the power and flexibility of human perception, ...
More
At the heart of human intelligence rests a fundamental puzzle: 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. This book 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, the book 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, the book presents unifying principles that govern human intelligence. First, inductive bias: any system that makes inferences based on limited data must constrain its hypotheses in some way before observing data. Second, approximation bias: 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, the book demonstrates that intelligent systems designed to meet these constraints yield characteristically human errors. Examining how humans make intelligent and maladaptive decisions, the book delves into the successes and failures of cognition.Less
At the heart of human intelligence rests a fundamental puzzle: 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. This book 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, the book 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, the book presents unifying principles that govern human intelligence. First, inductive bias: any system that makes inferences based on limited data must constrain its hypotheses in some way before observing data. Second, approximation bias: 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, the book demonstrates that intelligent systems designed to meet these constraints yield characteristically human errors. Examining how humans make intelligent and maladaptive decisions, the book delves into the successes and failures of cognition.
James C. G. Walker
- Published in print:
- 1991
- Published Online:
- November 2020
- ISBN:
- 9780195045208
- eISBN:
- 9780197560020
- Item type:
- chapter
- Publisher:
- Oxford University Press
- DOI:
- 10.1093/oso/9780195045208.003.0006
- Subject:
- Earth Sciences and Geography, Geochemistry
In a linear system the expressions, yp, for rates of change are linear functions of the dependent variables, y. More complicated functions of y do not appear, not even products of the dependent ...
More
In a linear system the expressions, yp, for rates of change are linear functions of the dependent variables, y. More complicated functions of y do not appear, not even products of the dependent variables like y1 * y2. But most theoretical problems in Earth system science involve nonlinearities. For example, the rate at which a chemical reaction consumes species 1 may be proportional to the product of the concentrations of species 1 and the species 2 with which it is reacting, y1 * y2. In this chapter I shall describe and solve a simple nonlinear system involving the reaction between dissolved oxygen and organic carbon in the deep sea. I shall show how the nonlinear system can be represented by a linear system, provided that changes in the dependent variables are made in sufficiently small increments. Such increments are kept small by stepping forward in time with small steps. The time step can be adjusted automatically during the calculation so as to keep the increments small enough but no smaller than necessary. Steps that are too large cause errors or even instability, and steps that are too small waste time. The representation of a nonlinear system by its linear equivalent (for small increments) calls on algebraic manipulations that can be tedious and a prolific source of mistakes in complicated systems. This algebra can be avoided, however, by letting the computer perform the equivalent manipulations numerically. I shall demonstrate how to do this, finishing the chapter with a program that can solve coupled nonlinear systems directly from the equations for the rates of change of the dependent variables, automatically adjusting the time step to small values when the rates of change are large and to large values when the rates of change are small. Figure 4-1 shows a simple model of the processes that control the oxygen balance of the deep sea. Dissolved oxygen in surface seawater equilibrates with the atmosphere. Downwelling currents carry dissolved oxygen into the deep sea, where most of it reacts (metabolically) with organic matter carried into the deep sea in the form of settling particles of biological origin.
Less
In a linear system the expressions, yp, for rates of change are linear functions of the dependent variables, y. More complicated functions of y do not appear, not even products of the dependent variables like y1 * y2. But most theoretical problems in Earth system science involve nonlinearities. For example, the rate at which a chemical reaction consumes species 1 may be proportional to the product of the concentrations of species 1 and the species 2 with which it is reacting, y1 * y2. In this chapter I shall describe and solve a simple nonlinear system involving the reaction between dissolved oxygen and organic carbon in the deep sea. I shall show how the nonlinear system can be represented by a linear system, provided that changes in the dependent variables are made in sufficiently small increments. Such increments are kept small by stepping forward in time with small steps. The time step can be adjusted automatically during the calculation so as to keep the increments small enough but no smaller than necessary. Steps that are too large cause errors or even instability, and steps that are too small waste time. The representation of a nonlinear system by its linear equivalent (for small increments) calls on algebraic manipulations that can be tedious and a prolific source of mistakes in complicated systems. This algebra can be avoided, however, by letting the computer perform the equivalent manipulations numerically. I shall demonstrate how to do this, finishing the chapter with a program that can solve coupled nonlinear systems directly from the equations for the rates of change of the dependent variables, automatically adjusting the time step to small values when the rates of change are large and to large values when the rates of change are small. Figure 4-1 shows a simple model of the processes that control the oxygen balance of the deep sea. Dissolved oxygen in surface seawater equilibrates with the atmosphere. Downwelling currents carry dissolved oxygen into the deep sea, where most of it reacts (metabolically) with organic matter carried into the deep sea in the form of settling particles of biological origin.
