Artificial Intelligence
| Aim: |
|
To introduce basic issues in AI.
|
| Plan: |
- Brief history of modern AI
- Some applications of AI
- Intelligent agents
- Symbolic and non-symbolic representations
|
What is Artificial Intelligence?
"AI is the boundary between what people can do and what computers can't [yet]."
History
Ancient Times
- Aristotle (Logic as an instrument for studying thought)
Early Efforts by Philsopher/Mathemticians
- Descartes (The mind body problem)
- Hume (Cognition is computation)
- Euler (1735 - representation of structure; search)
Formal Notation and Calculating Machines
- Leibnitz (1887 - first system of formal logic; calculating machines)
- Babbage (Programmable computing machines)
- Boole (1847, 1859)
- Frege (1879, 1884 - first-order predicate calculus)
Modern Founders of AI
- Alan Turing ("Computing Machinery and Intelligence"; Turing test)
- McCulloch & Pitts (neural nets)
- Norbert Wiener (cybernetics)
- John von Neumann (game theory)
- Claude Shannon (information theory)
- Newell & Simon (The Logic Theorist)
- John McCarthy (LISP, common sense reasoning)
- Marvin Minsky (Frames)
- Donald Michie (Freddy)
Achievements of AI
- Deep Thought is an international grand master chess player.
- Sphinx can recognise continuous speech without training for each speaker.
It operates in near real time using a vocabulary of 1000 words and has 94% word
accuracy.
- Navlab is a truck that can drive along a road at 55mph in normal traffic.
- Carlton and United Breweries use an AI planning system to plan production
of their beer.
- Robots are used regularly in manufacturing.
- Natural language interfaces to databases can be obtained on a PC.
- Machine Learning methods have been used to build expert systems.
- Expert systems are used regularly in finance, medicine, manufacturing, and
agriculture
How do we build an intelligent agent?
- Must be able to perceive its environment.
- Must be able to affect its environment.
- Must be able to reason about observations and actions
- Must be able to learn from observations and actions.
- Must have goals.
Symbolic Representations
To construct intelligent systems it is necessary to employ internal
representations of a symbolic nature, with cognitive activity corresponding to
computational manipulation of these symbolic representations. The symbolic
representations must refer to the external world.
Advantages of symbolic representation:
- The system builder can read what the system knows.
- Knowledge is represented by sentences in a formal language.
- It is possible to read the representation and understand the meaning of
the knowledge.
Non-symbolic Representations
Knowledge is represented by weights on connections in a network
Advantages of non-symbolic representation
- Can deal with combinations of attributes such as an image.
- Noise tolerant
CRICOS Provider Code No. 00098G