Download presentation
Presentation is loading. Please wait.
2
Introduction to Artificial Intelligence
CS311 Fall 2016 Ananya Christman
3
Course goals Be able to answer the question “What is AI?”
Learn and apply basic AI techniques… …to solve real-world problems, and in the process… …understand how HARD AI really is (and why)
4
AI is a huge field So, what is AI?
5
AI is a huge field So, what is AI? Many many definitions…
``The exciting new effort to make computers think ... machines with minds, in the full and literal sense'' (Haugeland, 1985) ``The automation of activities that we associate with human thinking, activities such as decision-making, problem solving, learning ...'' (Bellman, 1978) ``The study of mental faculties through the use of computational models'' (Charniak and McDermott, 1985) ``The study of the computations that make it possible to perceive, reason, and act'' (Winston, 1992) ``The art of creating machines that perform functions that require intelligence when performed by people'' (Kurzweil, 1990) ``The study of how to make computers do things at which, at the moment, people are better'' (Rich and Knight, 1991) ``A field of study that seeks to explain and emulate intelligent behavior in terms of computational processes'' (Schalkoff, 1990) ``The branch of computer science that is concerned with the automation of intelligent behavior'' (Luger and Stubblefield, 1993)
6
AI is a huge field So, what is AI?
One unifying definition: “Creating a machine that thinks (or acts) humanly (or rationally).” For example: reads, walks around, drives, plays games, solves problems, learns, has conversations… We will come back to this definition in tomorrow’s lecture. But for now let’s see how much we’ve accomplished in the field of AI.
7
HAL: from the movie 2001 2001: A Space Odyssey
classic science fiction movie from 1969 HAL “brains” of an intelligent spaceship speaks easily with the crew sees/understands emotions of the crew navigates the ship diagnoses on-board problems makes decisions displays emotions In 1969 this was science fiction: Is it still science fiction? HAL = Heuristically programmed Algorithmic Computer
8
What’s involved in building a computer like Hal?
HAL’s Legacy: 2001’s Computer as Dream and Reality MIT Press, 1997, David Stork (ed.) Speech interaction speech recognition speech understanding speaking ability Image recognition and understanding Learning Planning and Decision-Making
9
What’s involved in building a computer like Hal?
HAL’s Legacy: 2001’s Computer as Dream and Reality MIT Press, 1997, David Stork (ed.) Speech interaction speech recognition speech understanding speaking ability Image recognition and understanding Learning Planning and Decision-Making
10
Can Computers Speak? Loquendo (now Nuance) http://www.nuance.com/
Understandable, but may not always sound like a person
11
Can Computers Recognize Speech?
Map sounds from a mic into a list of words classic problem in AI, very difficult “Lets talk about how to wreck a nice beach” Easier?
12
Can Computers Recognize Speech?
Map sounds from a mic into a list of words classic problem in AI, very difficult “Lets talk about how to wreck a nice beach” Easier: recognizing single words from a set vocabulary Examples: Mac has built-in dictation software iPhone’s Siri Google allows you to search via voice command
13
Can Computers Understand Speech?
Understanding is different from recognition: “Time flies like an arrow” how many different interpretations are there? time passes quickly like an arrow command: time the flies the way an arrow would command: time only the flies that are like an arrow “time-flies” are fond of arrows
14
What’s involved in building a computer like Hal?
HAL’s Legacy: 2001’s Computer as Dream and Reality MIT Press, 1997, David Stork (ed.) Speech interaction speech recognition speech understanding speaking ability Image recognition and understanding Learning Planning and Decision-Making
15
Image Recognition Hard AI problem!
Humans can effortlessly recognize objects but computers can’t distinguish variations Some success with facial emotions: Kairos
16
What’s involved in building a computer like Hal?
HAL’s Legacy: 2001’s Computer as Dream and Reality MIT Press, 1997, David Stork (ed.) Speech interaction speech recognition speech understanding speaking ability Image recognition and understanding Learning Planning and Decision-Making
17
Learning Jeopardy Player: IBM’s Watson
18
Learning Jeopardy Player: IBM’s Watson Google’s Driverless Cars
Spam Filters Sentiment Analysis Rely heavily on training data
19
What’s involved in building a computer like Hal?
HAL’s Legacy: 2001’s Computer as Dream and Reality MIT Press, 1997, David Stork (ed.) Speech interaction speech recognition speech understanding speaking ability Image recognition and understanding Learning Planning and Decision-Making
20
Planning and Decision Making
Chess Player: IBM Deep Blue
21
Planning and Decision Making
Chess Player: IBM Deep Blue Navigational Robots Boston Dynamic’s Big Dog Roomba Cat vs. Roomba Google’s Driverless Cars again
22
HAL: from the movie 2001 Impossible? 2001: A Space Odyssey
classic science fiction movie from 1969 HAL “brains” of an intelligent spaceship speaks easily with the crew sees/understands emotions of the crew navigates the ship automatically diagnoses on-board problems makes decisions displays emotions In 1969 this was science fiction Is it still science fiction? Limited! Impossible? displays emotions
23
Course Info
24
How do we make a computer "smart?"
Reasoning with uncertainty Search Many different ways of making an agent intelligent Learning Reasoning with knowledge
25
What is an “agent”? “anything that can be viewed as perceiving its environment through sensors and acting upon that environment through actuators”
26
What is an “agent”? Living agent sensors = eyes, ears, etc
“anything that can be viewed as perceiving its environment through sensors and acting upon that environment through actuators” Living agent sensors = eyes, ears, etc actuators = hands, legs, mouth, etc Software agent sensors = any input devices – keyboard, camera, network, data files actuators = any output devices – screen, network, files (to write to)
27
Environments Fully observable vs. partially observable Deterministic vs. non-deterministic (stochastic) Benign vs. Adversarial Dynamic vs. Static
28
Roomba’s Environment
29
Environments Fully observable vs. partially observable
Do we have access to all of the relevant information? Deterministic vs. non-deterministic (stochastic) Is the outcome of an action always certain or is it probabilistic? Benign vs. Adversarial Does the environment try to oppose the agent? Dynamic vs. Static Does the environment change while the agent is deciding?
30
Roomba’s Environment Fully observable vs. partially observable
Do we have access to all of the relevant information? Deterministic vs. non-deterministic (stochastic) Is the outcome of an action always certain or is it probabilistic? Benign vs. Adversarial Does the environment try to oppose the agent? Dynamic vs. Static Does the environment change while the agent is deciding?
31
Roomba’s Environment Fully observable vs. partially observable
Do we have access to all of the relevant information? Deterministic vs. non-deterministic (stochastic) Is the outcome of an action always certain or is it probabilistic? Benign vs. Adversarial Does the environment try to oppose the agent? Dynamic vs. Static Does the environment change while the agent is deciding?
32
Roomba’s Environment Fully observable vs. partially observable
Do we have access to all of the relevant information? Deterministic vs. non-deterministic (stochastic) Is the outcome of an action always certain or is it probabilistic? Benign vs. Adversarial Does the environment try to oppose the agent? Dynamic vs. Static Does the environment change while the agent is deciding?
33
Roomba’s Environment Fully observable vs. partially observable
Do we have access to all of the relevant information? Deterministic vs. non-deterministic (stochastic) Is the outcome of an action always certain or is it probabilistic? Benign vs. Adversarial Does the environment try to oppose the agent? Dynamic vs. Static Does the environment change while the agent is deciding?
34
Roomba’s Environment Fully observable vs. partially observable
Do we have access to all of the relevant information? Deterministic vs. non-deterministic (stochastic) Is the outcome of an action always certain or is it probabilistic? Benign vs. Adversarial Does the environment try to oppose the agent? Dynamic vs. Static Does the environment change while the agent is deciding?
35
Roomba’s Environment Fully observable vs. partially observable
Do we have access to all of the relevant information? Deterministic vs. non-deterministic (stochastic) Is the outcome of an action always certain or is it probabilistic? Benign vs. Adversarial Does the environment try to oppose the agent? Dynamic vs. Static Does the environment change while the agent is deciding?
36
Roomba’s Environment Fully observable vs. partially observable
Do we have access to all of the relevant information? Deterministic vs. non-deterministic (stochastic) Is the outcome of an action always certain or is it probabilistic? Benign vs. Adversarial Does the environment try to oppose the agent? Dynamic vs. Static Does the environment change while the agent is deciding?
37
Roomba’s Environment Fully observable vs. partially observable
Do we have access to all of the relevant information? Deterministic vs. non-deterministic (stochastic) Is the outcome of an action always certain or is it probabilistic? Benign vs. Adversarial Does the environment try to oppose the agent? Dynamic vs. Static Does the environment change while the agent is deciding?
38
Back to Agents… Search Agents: Approach problem via search
To accomplish a task: Formulate problem and goal Search for a sequence of actions that will lead to the goal Execute the actions one at a time
39
Back to Agents… Search Agents: Approach problem via search
To accomplish a task: Formulate problem and goal Search for a sequence of actions that will lead to the goal Execute the actions one at a time
40
Formulating the problem:
What information does a search agent need to find a solution to a problem?
41
Formulating the problem for 8-Puzzle
What information does a search agent need to find a solution to a problem?
42
Formulating the problem:
Initial state: where do we start? what are all of the states? Actions: what are the possible actions? Transition model or state-space: given a current state, to which state does each applicable action take us? Goal/goal test: what is the end result we’re trying to achieve? Cost: what are the costs of the different actions?
43
Back to Agents… Search Agents: Approach problem via search
To accomplish a task: Formulate problem and goal Search for a sequence of actions that will lead to the goal Execute the actions one at a time
Similar presentations
