Artificial Intelligence A Brief Overview Of AI

Document Sample
Artificial Intelligence A Brief Overview Of AI Powered By Docstoc
					 Artificial Intelligence:
A Brief Overview Of AI
      Michael J. Sepcot
      Justin Brinkerhoff
                   History
Alan Turing
   Founder Of Computer Science
   Computing Machinery and Intelligence
   Turing Test
“Artificial Intelligence”
   1956: Dartmouth College
Intelligence
   “The ability to use optimally limited resources
    – including time – to achieve a set of goals.”
          Two Approaches
Mainstream
   Programs that teach themselves
   Recursive Algorithms
Neural Networks
   Learn from examples
   Simulation of human neurons
   Reverse engineer the human brain
      Programming Approach
LISP
   List Processing
   Developed in the 1950s
   Code developed by John McCarthy at MIT
AIAI – Artificial Intelligence Applications
Institute
   Computers operate themselves with little to
    no human input
      Programming Approach
Cyc – Pronounced “Psych”
   Commonsense approach to AI
AI Effect
   If a program fails, AI is not plausible. If a
    program succeeds, it was the programmer
    and not the machine; the machine was not
    that intelligent after all.
          Applications Of AI
General Problem Solver
Stock Market
Banking Systems
Gaming
   IBM Deep Blue
      Defeated Chess master Garry Kasparov
      Was a self-taught machine
            Applications Of AI
Medicine
   MYCIN
      An “Expert System”
      A system for evaluating complex cases involving
      meningitis.
      Diagnoses correct 65% of the time
          Doctor’s correct only 42.5 – 62.5% of the time
   BrainMaker
      Able to quickly and efficiently identify heart attacks
      due to enzyme data.
      Able to classify cancer cells.
                 Turing Test
    Definition - AI program attempting to pass as a
    human being in a conversation.
    The Test
     The test would be completed over computers with a
      human end and a machine end both communicating
      with a judge.
     Objectives
        The human judge would attempt to disclose which
        end user is the computer
        Questions to be asked can range from anything:
        military strategies to family life.
      Philosophical Issues
Are computers thinking, or are they just
calculating?

Once computers are as complex as the
human brain, and can match the human
brain in subtlety and complexity of thought,
are we to consider them conscious?
          Conclusion
History
Two Approaches
Programming
Applications
Intelligence