Transcript Week 3 - Computer Science - University of Birmingham

Introduction to AI
&
AI Principles (Semester 1)
WEEK 3 (07/08)
John Barnden
Professor of Artificial Intelligence
School of Computer Science
University of Birmingham, UK
Some Example Applications, contd. 2
(review)
Intelligent personalized web-search agents.
Policing and national security.
Machine translation (of language).
Semantic web.
WHAT DOES “ARTIFICIAL”
MEAN??
Links to Other Disciplines
Psychology: on how human and other animal minds and
behaviour work.
Physiology & Neurophysiology: on how bodies & brains
work.
Philosophy: on the nature of mind, thought,
consciousness, language, morality, imagination, reasoning,
representation, …
Linguistics: on the nature of communication via language
Sociology: on the nature and operation of society
others
How Do the Aims of AI Differ from those
of the Rest of CS?
 System ability on broader tasks or broader ranges of tasks;
Extrapolation to new variants of tasks.
 Adaptability, learning.
 Variable organization of major actions.
 Diversity of information.
 Handling of imperfections: uncertainty, vagueness, inconsistency,
incorrectness, missing info, system damage.
 “Graceful degradation” as info or system damage gets worse.
 “Satisficing”: being good enough rather than perfect.
 Self-reflection.
Expert versus Everyday/Common-Sense
Expert AI: e.g.,

Expert systems for medical diagnosis, etc.

Chess programs

Mathematical theorem provers

Aircraft movement planning systems

Specialized manufacturing robots
Everyday/Common-Sense AI: e.g.,

Language processing, for no specific domain

Common-sense reasoning about the everyday world

Seeing the everyday world
WHY IS EVERYDAY AI
CHALLENGING?
Who … … ?
How many of you … … ?
How many students at the University own cats?