Building Machines That Learn and Think Like People

Brenden Lake, Tomer D. Ullman, Joshua B. Tenenbaum, Samuel J. Gershman

Research output: Contribution to journalArticle

Abstract

Recent progress in artificial intelligence (AI) has renewed interest in building systems that learn and think like people. Many advances have come from using deep neural networks trained end-to-end in tasks such as object recognition, video games, and board games, achieving performance that equals or even beats humans in some respects. Despite their biological inspiration and performance achievements, these systems differ from human intelligence in crucial ways. We review progress in cognitive science suggesting that truly human-like learning and thinking machines will have to reach beyond current engineering trends in both what they learn, and how they learn it. Specifically, we argue that these machines should (a) build causal models of the world that support explanation and understanding, rather than merely solving pattern recognition problems; (b) ground learning in intuitive theories of physics and psychology, to support and enrich the knowledge that is learned; and (c) harness compositionality and learning-to-learn to rapidly acquire and generalize knowledge to new tasks and situations. We suggest concrete challenges and promising routes towards these goals that can combine the strengths of recent neural network advances with more structured cognitive models.

Original languageEnglish (US)
Pages (from-to)1-101
Number of pages101
JournalBehavioral and Brain Sciences
DOIs
StateAccepted/In press - Nov 24 2016

Fingerprint

neural network
Learning
learning
Cognitive Science
Video Games
pattern recognition
Artificial Intelligence
Physics
artificial intelligence
computer game
Intelligence
performance
physics
intelligence
building
psychology
engineering
Psychology
trend
science

ASJC Scopus subject areas

  • Neuropsychology and Physiological Psychology
  • Physiology
  • Language and Linguistics
  • Linguistics and Language
  • Behavioral Neuroscience

Cite this

Building Machines That Learn and Think Like People. / Lake, Brenden; Ullman, Tomer D.; Tenenbaum, Joshua B.; Gershman, Samuel J.

In: Behavioral and Brain Sciences, 24.11.2016, p. 1-101.

Research output: Contribution to journalArticle

Lake, Brenden ; Ullman, Tomer D. ; Tenenbaum, Joshua B. ; Gershman, Samuel J. / Building Machines That Learn and Think Like People. In: Behavioral and Brain Sciences. 2016 ; pp. 1-101.
@article{56ff9f5d9940497d85479208b217322c,
title = "Building Machines That Learn and Think Like People",
abstract = "Recent progress in artificial intelligence (AI) has renewed interest in building systems that learn and think like people. Many advances have come from using deep neural networks trained end-to-end in tasks such as object recognition, video games, and board games, achieving performance that equals or even beats humans in some respects. Despite their biological inspiration and performance achievements, these systems differ from human intelligence in crucial ways. We review progress in cognitive science suggesting that truly human-like learning and thinking machines will have to reach beyond current engineering trends in both what they learn, and how they learn it. Specifically, we argue that these machines should (a) build causal models of the world that support explanation and understanding, rather than merely solving pattern recognition problems; (b) ground learning in intuitive theories of physics and psychology, to support and enrich the knowledge that is learned; and (c) harness compositionality and learning-to-learn to rapidly acquire and generalize knowledge to new tasks and situations. We suggest concrete challenges and promising routes towards these goals that can combine the strengths of recent neural network advances with more structured cognitive models.",
author = "Brenden Lake and Ullman, {Tomer D.} and Tenenbaum, {Joshua B.} and Gershman, {Samuel J.}",
year = "2016",
month = "11",
day = "24",
doi = "10.1017/S0140525X16001837",
language = "English (US)",
pages = "1--101",
journal = "Behavioral and Brain Sciences",
issn = "0140-525X",
publisher = "Cambridge University Press",

}

TY - JOUR

T1 - Building Machines That Learn and Think Like People

AU - Lake, Brenden

AU - Ullman, Tomer D.

AU - Tenenbaum, Joshua B.

AU - Gershman, Samuel J.

PY - 2016/11/24

Y1 - 2016/11/24

N2 - Recent progress in artificial intelligence (AI) has renewed interest in building systems that learn and think like people. Many advances have come from using deep neural networks trained end-to-end in tasks such as object recognition, video games, and board games, achieving performance that equals or even beats humans in some respects. Despite their biological inspiration and performance achievements, these systems differ from human intelligence in crucial ways. We review progress in cognitive science suggesting that truly human-like learning and thinking machines will have to reach beyond current engineering trends in both what they learn, and how they learn it. Specifically, we argue that these machines should (a) build causal models of the world that support explanation and understanding, rather than merely solving pattern recognition problems; (b) ground learning in intuitive theories of physics and psychology, to support and enrich the knowledge that is learned; and (c) harness compositionality and learning-to-learn to rapidly acquire and generalize knowledge to new tasks and situations. We suggest concrete challenges and promising routes towards these goals that can combine the strengths of recent neural network advances with more structured cognitive models.

AB - Recent progress in artificial intelligence (AI) has renewed interest in building systems that learn and think like people. Many advances have come from using deep neural networks trained end-to-end in tasks such as object recognition, video games, and board games, achieving performance that equals or even beats humans in some respects. Despite their biological inspiration and performance achievements, these systems differ from human intelligence in crucial ways. We review progress in cognitive science suggesting that truly human-like learning and thinking machines will have to reach beyond current engineering trends in both what they learn, and how they learn it. Specifically, we argue that these machines should (a) build causal models of the world that support explanation and understanding, rather than merely solving pattern recognition problems; (b) ground learning in intuitive theories of physics and psychology, to support and enrich the knowledge that is learned; and (c) harness compositionality and learning-to-learn to rapidly acquire and generalize knowledge to new tasks and situations. We suggest concrete challenges and promising routes towards these goals that can combine the strengths of recent neural network advances with more structured cognitive models.

UR - http://www.scopus.com/inward/record.url?scp=84996868095&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84996868095&partnerID=8YFLogxK

U2 - 10.1017/S0140525X16001837

DO - 10.1017/S0140525X16001837

M3 - Article

SP - 1

EP - 101

JO - Behavioral and Brain Sciences

JF - Behavioral and Brain Sciences

SN - 0140-525X

ER -