Logic, methods of reasoning, mind as physical system foundations of learning, language, rationality
ì
Mathematics
Formal representation and proof algorithms, computation, (un)decidability, (in)tractability, probability
ì
Economics
utility, decision theory
ì
Neuroscience
physical substrate for mental activity
ì
Psychology
cognitive science, affective science
ì
Computer engineering
efficient algorithms
ì
Control theory
design systems that maximize an objective function over time
ì
Linguistics
knowledge representation, grammar
1956: The Birth of AI …solve kinds of problems now reserved for humans…
…significant advance can be made in one or more of these problems if a carefully selected group of scientists work on it together for a summer…
Abridged history of AI ì ì ì ì ì
1943 1950 1956 1952—69 1950s
McCulloch & Pitts: Boolean circuit model of brain Turing's "Computing Machinery and Intelligence" Dartmouth meeting: "Artificial Intelligence" adopted Look, Ma, no hands! Early AI programs, including Samuel's checkers program, Newell & Simon's Logic Theorist, Gelernter's Geometry Engine
1960s: Initial Optimism ì Playing checkers
(Arthur Samuel)
ì General Problem Solver
(Allen Newell & Herbert Simon)
Abridged history of AI ì 1965 ì 1966—73 ì 1969—79
Robinson's complete algorithm for logical reasoning AI discovers computational complexity Neural network research almost disappears Early development of knowledge-based systems Blocks world SHRDLU (Winograd 1972)
Abridged history of AI
Abridged history of AI
Road block? ì What do you think the road block to these types of systems
were?
Abridged history of AI ì ì ì ì
1980-1986-1987-1995--
AI becomes an industry Neural networks return to popularity AI becomes a science The emergence of intelligent agents
1990s: Resurgence ì ProbabilisLc approaches ì AdopLon of scienLfic
rigor
ì Return to dreams of
human-‐level AI
2000s: Big Data ì Focus on data
ì SoluLon to KA
boPleneck?
What is AI?
ì
What is AI? Views of AI fall into four categories: Thinking humanly
Thinking raLonally
AcLng humanly
AcLng raLonally
The textbook advocates "acLng raLonally"
Thinking humanly: cognitive modeling ì CogniLve Modeling Approach ì 1960s "cogniLve revoluLon": informaLon-‐processing psychology ì Requires scienLfic theories of internal acLviLes of the
brain ì ValidaLon Requires
1) PredicLng and tesLng behavior of human subjects (top-‐down) or 2) Direct idenLficaLon from neurological data (boPom-‐up)
Systems modeled based on the way humans think. Systems are based on:
ì ProducLons ì Rule-‐based ì Experimental results of human behavior
Acting humanly: Turing Test ì Turing (1950) "Computing machinery and intelligence": ì "Can machines think?" à "Can machines behave
intelligently?”
Image from: Hwee Tou Ng’s AI lectures
Acting humanly: Turing Test ì Stimulated the development of: ì Natural Language Processing ì Knowledge RepresentaLon ì Automated Reasoning ì Machine Learning ì Computer Vision ì RoboLcs
Eliza
ELIZA Person : Men are all alike. ELIZA : In what way? Person : They're always bugging us about something or other. ELIZA : Can you think of a specific example? Person : Well, my boyfriend made me come here. ELIZA : Your boyfriend made you come here? Person : He says I'm depressed much of the Lme. ELIZA : I am sorry to hear that you are depressed. Person : It's true. I am unhappy. ELIZA : Do you think coming here will help you not to be unhappy? Person : I need some help, that much seems certain. ELIZA : What would it mean to you if you got some help? Person : Perhaps I could learn to get along with my mother. ELIZA : Tell me more about your family
Aristotle: what are correct arguments/ thought processes?
ì
Purely logical thought and reasoning
ì
MathemaLcally well-‐defined
ì
Problems: 1. 2.
Not all intelligent behavior is mediated by logical deliberaLon What is the purpose of thinking? What thoughts should I have?
Acting rationally: rational agent Agent: entity that perceives and acts Rational behavior: choose behavior that maximize goal achievement, given the available information
Agent-‐Based AI
Agent Design ì
What can the agent do? ì
ì
What is the environment? (Input: percepts) ì
ì
How is it interpreted?
What does the agent know? ì ì ì ì
ì
Range of ac5ons
History of previous inputs and acLons (how far back?) ProperLes of environment: world knowledge Knowledge of its own goals and preferences Strategies for behavior
How does the agent choose to act? ì
Mapping from percept sequence -‐> acLon called an agent func5on
