Download Introduction to Artificial Intelligence

Survey
yes no Was this document useful for you?
   Thank you for your participation!

* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project

Document related concepts

Agent-based model wikipedia , lookup

Machine learning wikipedia , lookup

The City and the Stars wikipedia , lookup

Technological singularity wikipedia , lookup

AI winter wikipedia , lookup

Intelligence explosion wikipedia , lookup

Agent (The Matrix) wikipedia , lookup

Existential risk from artificial general intelligence wikipedia , lookup

Embodied cognitive science wikipedia , lookup

Philosophy of artificial intelligence wikipedia , lookup

Ethics of artificial intelligence wikipedia , lookup

Collaborative information seeking wikipedia , lookup

History of artificial intelligence wikipedia , lookup

Transcript
Introduction to Artificial Intelligence
Mitch Marcus
CIS391
Fall, 2015
Welcome to CIS 391
 Lecturer:
• Mitch Marcus, mitch@<cis standard>
• Levine 503
• Office hours will be announced on Piazza
 TA:
• Daniel Moroz, dmoroz@<seas standard>
• Office hours will be announced on Piazza
 Course Administrator:
Cheryl Hickey, cherylh@<seas standard>
Levine 502,
215-898-3538
CIS 391 - Intro to AI - Fall 2015
2
Welcome to CIS 391
 Course home page: http://www.seas.upenn.edu/~cis391
 Discussion via Piazza (link on course home page)
 Textbook: S. Russell and P. Norvig Artificial Intelligence: A
Modern Approach Prentice Hall, 2009, Third Edition (U.S.)
 Prerequisites:
• CIS 120, 121 & 160 (Not CIS 262)
• Introductory probability and statistics will be very useful
• Familiarity with propositional logic and finite state automata will be
useful.
• We assume ability to master Python after a couple of class lectures.
CIS 391 - Intro to AI - Fall 2015
3
Grading & Homework
Grading:



50% Homeworks
25% Midterm 1
25% Midterm 2
Homework:





Homework will be due at 11:59 on specified dates with submission cut off
promptly.
You can submit up to two homeworks late, but extensions after that will be
granted only for true emergencies.
Your lowest homework grade will be dropped.
ALL HOMEWORKS MUST BE YOUR OWN INDEPENDENT WORK
Violations of Penn's Code of Academic Integrity and in particular
academic dishonesty as defined in the Code of Integrity will not be
tolerated. PENALTIES WILL BE SUBSTANTIAL
CIS 391 - Intro to AI - Fall 2015
4
On to the Real Stuff:
WHAT THE COURSE IS AND
ISN’T ABOUT
CIS 391 - Intro to AI - Fall 2015
5
“I want to design a machine that will
be proud of me – Danny Hillis”
CIS 391 - Intro to AI - Fall 2015
6
“I want to design a machine that will
be proud of me – Danny Hillis”
CIS 391 - Intro to AI - Fall 2015
7
“Startup Funded $143M to Create Sentient
Computing” – EETimes 12/2014

Now a startup with $143 million in funding [is] describing a sentient
distributed artificial intelligence that sounds like a nice-guy version of
Skynet from the cinema flick Terminator. According to the technology
gurus at Sentient Technologies Holdings Ltd. of San Francisco, the
software for sentient computers, which they are already installing at key
customer sites, goes beyond natural language recognition, unstructured
searching, machine learning, and deep knowledge.

"Reasoning and logic are one thing,
but beyond that is true intelligence
-- what we call sentience," Babak
Hojat, cofounder and chief scientist
tells EE Times.
"Sentience is being aware, having
perceptions, being mindful, and has
implications of autonomy," chief
technology officer Nigel Duffy said.

CIS 391 - Intro to AI - Fall 2015
8
“Startup Funded $143M to Create Sentient
Computing” – EETimes 12/2014

Now a startup with $143 million in funding [is] describing a sentient
distributed artificial intelligence that sounds like a nice-guy version of
Skynet from the cinema flick Terminator. According to the technology
gurus at Sentient Technologies Holdings Ltd. of San Francisco, the
software for sentient computers, which they are already installing at key
customer sites, goes beyond natural language recognition, unstructured
searching, machine learning, and deep knowledge.

"Reasoning and logic are one thing,
but beyond that is true intelligence
-- what we call sentience," Babak
Hojat, cofounder and chief scientist
tells EE Times.
"Sentience is being aware, having
perceptions, being mindful, and has
implications of autonomy," chief
technology officer Nigel Duffy said.

CIS 391 - Intro to AI - Fall 2015
9
Recent Significant Advances In NLP
 IBM’s Watson
 Web-scale information extraction
& question answering
 Apple’s Siri
 Interactive Dialogue Systems
 Google Translate
 Automatic Machine Translation
CIS 391 - Intro to AI - Fall 2015
10
Broadcast Monitoring
BBN MAPS & Language Weaver MT
CIS 391 - Intro to AI - Fall 2015
11
CIS 391 - Intro to AI - Fall 2015
12
A REAL Achievement : DARPA Grand Challenge 2005
CIS 391 - Intro to AI - Fall 2015
13
DARPA Urban Challenge 2007
CIS 391 - Intro to AI - Fall 2015
14
Older Real Accomplishments of AI
 1991: AI Logistics Planning for Gulf War
 1997: Deep Blue defeated the reigning world chess
champion Garry Kasparov
 1998: Deep Space 1 (launched) – Remote Agent Experiment
“Invisible” AI
 Computer Algebra Systems (Maple, Mathematica)
 Machine Learning
• Credit Evaluation, Fraud Detection
• Internet Search, Spam Filtering
• Handwritten character recognition (checks, US mail)
CIS 391 - Intro to AI - Fall 2015
15
Thinking humanly Thinking rationally
Acting humanly
What is AI?
Acting rationally
Views of AI fall into four categories:
Thinking humanly
Thinking rationally
Acting humanly
Acting rationally
We will focus on "acting rationally“
CIS 391 - Intro to AI - Fall 2015
16
Acting humanly:
Turing Test
Thinking humanly Thinking rationally
Acting humanly
Acting rationally

Turing (1950) "Computing machinery and intelligence":
“Can machines think?”  “Can machines behave intelligently?”

Operational test for intelligent behavior: the Imitation Game

Predicted that by 2000, a machine might have a 30% chance of
fooling a lay person for 5 minutes

Anticipated most major arguments against AI

Suggested major components of AI: knowledge, reasoning,
language understanding, learning
CIS 391 - Intro to AI - Fall 2015
17
Acting humanly:
Social robots
Thinking humanly Thinking rationally
Acting humanly
Acting rationally
 Cynthia Breazeal: MIT
CIS 391 - Intro to AI - Fall 2015
18
Thinking humanly:
Thinking humanly Thinking rationally
Acting humanly
Acting rationally
cognitive modeling
 1960s "cognitive revolution": informationprocessing psychology, a.k.a. cognitive psychology
 Requires scientific theories of internal activities of the brain

How to validate? Requires
1) Predicting and testing behavior of human subjects
or
2) Direct identification from neurological data (bottom-up) :
Cognitive Neuroscience
 Both approaches are now distinct from AI
 Caveat: ACT-R & SOAR communities do computational
modeling of high level mental functions
CIS 391 - Intro to AI - Fall 2015
19
Thinking rationally:
"laws of thought"
Thinking humanly Thinking rationally
Acting humanly
Acting rationally

Aristotle: what are correct arguments/thought processes?

Several Greek schools developed various forms of logic:
notation and rules of derivation for thoughts; may or may
not have proceeded to the idea of mechanization

Direct line through mathematics and philosophy to
modern AI

Problems:
1.
2.
3.
4.
5.
Not all intelligent behavior is mediated by logical deliberation
What is the purpose of thinking? What thoughts should I have?
Ignores the hard problem of perception
All attempts to encode what we know in logic have failed
Most logical inference is intractable
CIS 391 - Intro to AI - Fall 2015
20
Acting rationally:
rational agents
Thinking humanly Thinking rationally
Acting humanly
Acting rationally
 Rational behavior: doing the right thing
 The right thing: that which is expected to
maximize goal achievement, given the available
information
 Doesn't necessarily involve thinking – e.g.,
blinking reflex – but thinking should be in the
service of rational action
CIS 391 - Intro to AI - Fall 2015
21
Thinking humanly Thinking rationally
Rational agents
Acting humanly
Acting rationally
Rational agent: An agent is an entity that
perceives and acts
This course is about effective
programming techniques for
designing rational agents
CIS 391 - Intro to AI - Fall 2015
22
Agents and environments
An agent is specified by an agent function f:P a
that maps sequences of percept vectors P to an
action a from a set A:
P=[p0, p1, … , pt]
A={a0, a1, … , ak}
CIS 521 - Intro to AI - Winter 2015
23
Agents
 An agent is anything that can be viewed as
• perceiving its environment through sensors and
• acting upon that environment through actuators
 Human agent:
• Sensors: eyes, ears, ...
• Actuators: hands, legs, mouth, …
 Robotic agent:
• Sensors: cameras and infrared range finders
• Actuators: various motors
 Agents include humans, robots, softbots, thermostats, …
CIS 391 - Intro to AI - Fall 2015
24
Agent function & program
 The agent program runs on the physical
architecture to produce f
• agent = architecture + program
 “Easy” solution: table that maps every possible
sequence Y to an action a
• One small problem: exponential in length of Y
CIS 391 - Intro to AI - Fall 2015
25
Rational agents II
 Rational Agent: For each possible percept
sequence, a rational agent should select an
action that is expected to maximize its
performance measure.
 Performance measure: An objective criterion for
success of an agent's behavior, given the
evidence provided by the percept sequence.
 A performance measure for a vacuum-cleaner
agent might include one or more of:
• +1 point for each clean square in time T
• +1 point for clean square, -1 for each move
• -1000 for more than k dirty squares
CIS 391 - Intro to AI - Fall 2015
26
Rationality is not omniscience
 Ideal agent: maximizes actual performance, but
needs to be omniscient.
• Usually impossible…..
— But consider tic-tac-toe agent…
• Rationality  Success
 Agents can perform actions in order to modify
future percepts so as to obtain useful information
(information gathering, exploration)
 Caveat: computational limitations make perfect
rationality unachievable
 design best program for given machine resources
CIS 521 - Intro to AI - Winter 2015
27
Two Approaches to AI
 Logical representations: BEFORE 1995
• Relations between entities
—“Mitch’s bicycle is red”
– (isa B3241 bicycle) (color B3231 red) (owns B3241 P119)
– (isa P119 person) (name P119 “Mitch”)
• Explicit logical models
• Logical inference, Search
• Chess, Sudoko, computer games, …
 Statistical models: SINCE 2000
• Prediction by look-up or by weighted combinations
—P(y=bicycle) = c0 + c1 x1 +c2 x2 + c3 x3 + …
• Machine Learning, Machine vision, speech recognition, …
CIS 391 - Intro to AI - Fall 2015
28
Course Overview – First Half
Module 0: Introduction
• Intelligent Agents
• Python Programming
Module 1: Search Strategies
• Uninformed & Informed Search (Homeworks: Puzzle Solvers)
• Constraint Satisfaction (Homework: Sudoku Solver)
• Adversarial Search
CIS 391 - Intro to AI - Fall 2015
29
Course Overview – Second Half
Module 2: Machine Learning and Natural Language
Processing
• Review of Probability
• Naive Bayes (Spam Filtering) & Bayesian Networks
(Homework: Build a spam filter)
• Perceptrons and Support Vector Machines
• Hidden Markov Models & Part of Speech Tagging
(Homework: Generate fake Frankenstein text, Build Part of
Speech Tagger)
Module 3: Knowledge Representation and Logic
• Logical Agents (Homework: Logic Puzzle Solver)
• The Singularity: A critique
CIS 391 - Intro to AI - Fall 2015
30
The last lecture: Kurweil’s singularity vision
CSE 391 - Into to AI
31
The last lecture: Kurweil’s singularity vision
“With artificial intelligence we’re
summoning the demon”
– Elon Musk
“Full artificial intelligence could
spell the end of the human race”
– Steven Hawking
CSE 391 - Into to AI
32