Transcript File

Outline
•
•
•
•
•
•
Administrative issues
Course overview
What are Intelligent Systems?
A brief history
State of the art
Intelligent agents
Course overview
I. Introduction IS
II. Intelligent agents
III. Search
IV. Knowledge and reasoning
V. Planning
VI. Uncertainty
VII. Learning
VIII.Hybrid systems
Course overview
I. Introduction IS
II. Intelligent agents
III. Search
IV. Knowledge and reasoning
V. Planning
VI. Uncertainty
VII. Learning
VIII.Hybrid systems
What is an Intelligent System?
•System that:
•does everything that I want it to do & nothing that
I don’t
•Human-like reasoning
•recover from failure
•Learns from its mistakes
•Adapts to its environment
•Makes decision about appropriateness of actions
What are Intelligent Systems?
(What is AI?)
Turing test
Logic
Systems that think
like humans
Systems that think
rationally
Systems that act
like humans
Systems that act
rationally
Cognitive
science
Agents
Turing test (‘50)
Turing test pro’s & cons
• Predicted that by 2000 a machine might have a
30% chance of fooling a lay person for 5 min
• Anticipated major arguments against AI
– Pb: not reproducible, constructive, allows no
mathematical analysis
• Suggested major components of AI: knowledge,
reasoning, language understanding, learning
Acting rationally
• Doing the right thing
• Expected to maximize goal achievement,
given available info
• Doesn’t necessarily involve thinking!!
– Reflexes, etc.
Dimensions of AI
1.
2.
3.
4.
5.
symbolic or sub-symbolic (i.e., numerical)
manually engineered or automatically learned
cognitively plausible
perfect world assumption or probabilistic
evaluated by hunch, single example, authoritative tone
of voice or numerical measure
6. classification, numerical or planning
7. full system or human-assistor
8. online or offline
9. real-time or offline
10. intelligent or just software engineering
11. sound theoretical foundation or trial&error
(some) Categories of intelligent
system software
All software
Expert
systems
NN
Bayesian
Evolutionary
updating,
alg.
Certainty theory,
Fuzzy logic
Simulated
Objects,
Rule based
Annealing
frames,
systems
agents Knowledge
based
Computational
systems
intelligence
Example: intelligent agent Steve
http://www.isi.edu/isd/VET/steve-demo.html
Acting rationally + humanly
Collaboration
Example: evolution
Evolving artificial creatures, Karl Sims:
http://biota.org/ksims/blockies/index.html#video
Other Intelligent Systems
Distributed intelligence
Applications of Intelligent
Systems
Applications of Adaptive
Systems
• expert systems
– (e.g. medical diagnosis)
• data mining
– (e.g. search engines)
• computational linguistics
• games
More Applications
• Parallel computing:
– evolution of cellular automata
• Molecular biology:
– molecular evolution, design of useful molecules,
protein design
• Computer security:
– immune systems for computers
• Intelligent agents and robotics – Internet!!
• Scientific modeling:
– evolution, ecologies, economies, insect societies,
immune systems, organizations
AI is New Science
• We're still learning about perceiving,
planning and learning
• It is hard to compare methods/paradigms
You'll be pointed at research + literature
I'm you're tour guide: methods and
applications
Should we replace people with
machines?
• Yes. Expand the meaning of "grunt-work", and make a
machine do it.
• No. The devastating dehumanisation of automation.
• No. Economic objections?
• No. A National Aeronautics and Space Administration
report supporting manned space exploration states, "Man
is the lowest cost, 150-pound, non-linear, all-purpose
computer system that can be mass produced by unskilled
labour."
– with built-in sense-of-humour system
• Yes. what about the expense of life support and litigation?
• No. Nothing's gonna replace people; I'm a humanitarian
Course overview
I. Introduction IS
II. Intelligent agents
III. Search
IV. Knowledge and reasoning
V. Planning
VI. Uncertainty
VII. Learning
VIII.Hybrid systems
II. Intelligent agents
•
•
•
•
•
Rational agent: generalities
Agent & its environment
Example: a simple agent
Rationality defined
Task environment:
– PEAS
– Proprieties
• Agent proprieties