Situation Calculus

1
TDT 4136 Logikk og Resonnerende Systemer
http//www.idi.ntnu.no/emner/tdt4136/
http//www.idi.ntnu.no/toreamb/ JAVA/Newlogo.html
2
Overall course structure
AI Artificial Intelligence (A modern approach)
AI-2 Spring semester TDT4171 Methods in
artificial intelligence
AI-1 Fall semester TDT 4136 Logic and reasoning
systems
3
Course prehistory
Before
IT2702 Artificial Intelligence
TDT4170 Knowledge systems
TDT 4125 Logic
Now
TDT4136 Logic and reasoning systems
TDT4171 Methods in artificial intelligence
4
What is intelligence ?
Intelligence is the ability to solve new problems
based on earlier experience.
5
(No Transcript)
6
(No Transcript)
7
(No Transcript)
8
(No Transcript)
9
(No Transcript)
10
SICStus Prolog
Prolog Compiler developed by Swedish Institue
of Computer Science (SICS) 1990 (?)
Performance on a fast PC (3.4 GHz) gt 70
MegaLIPS (70 Mill. Logic inferences/second)
11
What is Artificial Intelligence
Artificial Intelligence is the science of making
machines do things that would require
intelligence if done by man.
(Prof. Marvin Minsky)
AI is the study of how to make computers do
things which at the moment people do better.
(Prof.
Elaine Rich) (Horizon effect If it
works, it is no longer AI)
12
A Definition of Intelligence
An entity is intelligent if it has an adequate
model of the world, it is clever enough to answer
a wide variety of questions on the basis of this
model, if it can get additional information from
the external world when required, and can perform
such tasks in the external world as its goals
demands and physical abilities permit.
13
What is AI (N.J.Nilsson)
AI is concerned with intelligent
behaviour. Intelligent behaviour involves
perception, reasoning, learning, communicating
and acting in complex environments. AI has as
one of its long term goals the development of
machines that can do these things as well as
humans can. Another goal is to understand this
kind of behaviour. Thus, AI has both engineering
and scientific goals.
14
Artificial Intelligence (AI)
Grand goal is to achieve human level intelligence.
AI was coined at the Dartmouth conference in
1956. Founding father John McCarthy.
15
(No Transcript)
16
What is Artificial
TRACTOR
ARTIFICIAL INTELLIGENCE
Artificial horse ?
INTELLIGENCE
HORSE
17
The Intelligence Pyramid
Each level is a set of relations on the level
below
Wisdom
Intelligence
Knowledge
Information
Data
Noise
18
Levels of intelligence
7. Can A be alive (not allowed to be killed)
? 6. Can A have a (genuine) consciousness ? 5.
Can A feel real feelings (pain,sorrow,happiness)
? 4. Can A think ? 3. Can A reason ? 2. Can A
deduce ? 1. Can A compute ?
Replace A with human/child/embryo/ape/robot/compu
ter What is your opinion ?
19
(No Transcript)
20
Aspects of Intelligence
Human likeness (how much does it
resemble a human ?) Human performance (is it
as clever as a human ?) Human like tasking
(are humans doing it ?) Human like operation
(are humans doing it similarily?) Human like
genesis (was it made or did it evolve?)
21
(No Transcript)
22
(No Transcript)
23
(No Transcript)
24
What is AI (James Allen,99)

• AI is the science of making machines that do
  tasks that humans can do or try to do
• AI is not the science of building artificial
  people
• AI is not the science of understanding human
  intelligence (Cognitive Science)
• AI is not even the science of trying to build
  artifacts that can imitate human behaviours well
  enough to fool someone that the machine is human
  (Turings test)

25
(No Transcript)
26
Game Playing
Om May 11 1997, an IBM program DEEP BLUE beat the
reigning world champion Garry Kasparov 3.5 2.5
in a six game match.
We won ?
27
(No Transcript)
28
(No Transcript)
29
The Noble Art of Car Parking
30
(No Transcript)
31
(No Transcript)
32
(No Transcript)
33
Artificial Expertise
Expertise (50 cents)
Expertise (50 cents
If most people prefer the computer to the human
expert, the computer has artificial expertise.
34
Rule Based Systems
Rule based systems / Knowledge based systems/
Expert Systems have played and plays an
important role in the AI industry. A report from
from 1993 by John Durkin Reports on Over 2500
Developed Expert Systems Application
areas Agriculture, Business, Chemistry,
Communications, Computer Systems, Education,
Electronics, Engineering, Environment, Geology,
Image processing, Information Management, Law,
Manufacturing, Mathematics, Medicine,
Meteorology, Military, Mining, Power Systems,
Science, Space Technology, Transportation Types
of systems Rule Based, Frame Based, Fuzzy
Logic, Case Based, Neural Network
35
Architecture of a typical expert system
Knowledge base
User interface Question-and-answer Menu
driven Natural language Graphic inteface
Knowledge- base editor
General knowledge- base
Inference engine
User
Case-specific data
Explanation subsystem
Expert system shell
36
AI in Medicine (USA 1970)

• Stanford
• MYCIN - blood infections
• Rutgers
• CASNET - casual reasoning
• MIT
• PIP - renal disease
• Stanford
• Pittsburgh
• Internist internal medicine

- the primary goal of this field is to develop
computer programs that perform efficiently and
are able to explain their reasoning and
conclusions to their users
37
Mycin system for diagnosis og meningitis and
bacteremia (bacterial infections)
IF the site of the culture is blood, and
the identity of the organism is not
known with certainty, and the stain of the
organism is gramneg, and the morphology of the
organism is rod, and the patient has been
seriously burned THEN there is weakly
suggestive evidence (0.4) that the
identity of the organism is pseudomonas
38
Intelligent Programming/ Programmed Intelligence
For a given task, it is possible to make a
program that performs intelligently (Intelligent
programming) . Then the intelligence is implicit
in the program.
A goal is to make a program for any given set of
tasks can perform intelligently without being
reprogrammed. (Programmed Intelligence). This
requires that the intelligence is represented
explicitly.

• Machine Learning
• Genetic Algorithms
• Evolutinary Programming

39
(No Transcript)
40
The Bus Route Oracle BussTUC
BussTUC is a natural language expert system for
bus departures in Trondheim. It is written in
Prolog ( gt 110.000 program lines) and answers gt
800.000 queries/year (2007)
41
(No Transcript)
42
(No Transcript)
43
Game of Life, Example of Artificial life
http//www.idi.ntnu.no/toreamb/JAVA/Babylon5.html
44
Artificial Intelligence versus Cognitive Science
Although a computer can do logical reasoning, it
does not mean that the computer is trying to
simulate a human. In fact, computers can do
logical reasoning better than humans. We can
say, with a twist, that AI is the science of
correct thinking, while CS is the science of
incorrect thinking. (Errare humane est).
45
Tower of Hanoi Puzzle
Cognitive Science How do humans solve the TOH
problem? Artificiel Intelligence How can we
make the machine solve it efficiently and
autonomously?
46
Are we machines ? Can silicon computers think ?
If humans are machines, then machines can think.
Even if machines made of proteins can think,
perhaps ones made of silicon does not. (Searle,
1992) Chinese room scenario
47
What is understanding ?Searles Chinese Room
Does the system understand Chinese ? Is the
system conscious ? What if John has the rules in
his head ?
? John
Rules
48
Physical Symbol System Hypothesis
That hypothesis states that a physical symbol
system has the necessary and sufficient means
for general intelligent action. (NewellSimon,
1976) A physical symbol system is a machine,
like a digital computer that is capable of
manipulating symbolic data. It doesnt matter
what the physical symbol system is made of.
49
(No Transcript)