introduction to artificial intelligence cs 438 spring 2008 today –aima, chapter 1 –defining ai...
TRANSCRIPT
Introduction to Artificial Intelligence
CS 438 Spring 2008• Today
– AIMA, Chapter 1– Defining AI
• Next Tuesday– Intelligent Agents– AIMA, Chapter 2– HW: Problem 2.5
• ADD: “A vacuum cleaning robot”
Goals of Artificial Intelligence
Human
Performance
Ideal
Performance
Thinking Systems that think like human
Cognitive Science
Systems that think rationally
Formal Logic
Behavior Systems that act like humans
Turing Test
Systems that act rationally
Heuristic Reasoning
Thinking Humanly• Cognitive Science
– Construct theories of how the human mind works
Thinking Humanly
• The Game of 15
1. Two players alternate in picking numbers from 1 to 9
2. Each number may only be used once
3. The player who picks 3 numbers that add up to exactly 15 wins
Thinking Humanly
• Nine-cell magic square is equivalent to Tic-Tac-Toe
• Same game, but magic square representation is easier for a computer to reason with, Tic-Tac-Toe Board is easier for a human to reason with
Thinking Rationally
• Rationality– Thinking “right” or “correctly”
• Thinking Rationally– Logic: formal laws of
thought• Patterns of representing
knowledge that guarantee every inference is correct
Thinking Rationally
• Deduction– Aristotle is a man – All men are mortal– Therefore: Aristotle is mortal
• Induction– Aristotle is a man– Aristotle is mortal– Therefore: All men are mortal
• Abduction– All men are mortal– Aristotle is mortal– Therefore: Aristotle is a man
Thinking Rationally
• Problems as a KR– Informal knowledge is difficult to state in
formal terms– Small problems can quickly exhaust
computational resources– Not always a provably correct thing to do
Acting Humanly
• Turing Test– An operational definition of intelligence– Can a computer carry on a conversation well enough
to convince you that you are talking to another human– Loebner Prize– What is needed to pass the TT
• NLP (Natural Language Processing)• KR • Automated Reasoning• Machine Learning
Total Turing Test
• Computer Vision (or other senses)• Computer manipulation• Robot
– A machine able to extract information from its environment and use knowledge about its world to move safely in a meaningful manner
– Physically embodied Intelligent Agent• As opposed to a “brain-in-a-box”
What does it take to get an intelligent robot to do a simple task?
Robot Parts: Two Arms, Vision, and Brain
The Brain can communicate with all parts
Arms can take commands as left, right, up, down, forward, and backward
Arms can answer yes/no about whether they are touching something but cannot distinguish what they are touching
The vision system can answer any question the brain asks, but cannot volunteer information.
The vision system can move around to get a better view.
Why is this simple task so difficult?
• Coordination is difficult
• Indirect feedback
• Updating knowledge about the environment
• Unexpected events– Need to re-plan
• Different coordinate systems need to be resolved– Box-centered and arm-centered
Acting Rationally• Acting to achieve one’s goals given one’s beliefs
– Given the information it knows at the time and its knowledge base it performs THE BEST action to achieve its goals
– Uncertainty• Making the action with the best expected outcome
– Bounded rationality• Determining the best action with limitations of resources of
time and/or memory
• AI: The construction of rational agents
Physical Symbol System hypothesis
• "A physical symbol system has the necessary and sufficient means of general intelligent action.“ A. Newell & H. Simon
Intelligent Agents (IA’s) are Physical Symbol Systems– Symbols and symbols structures that can be
manipulated syntactically by a set of processes– The symbol structure can be interpreted
semantically