418 01 context for ai


Published on

  • Be the first to comment

  • Be the first to like this

No Downloads
Total views
On SlideShare
From Embeds
Number of Embeds
Embeds 0
No embeds

No notes for slide
  • Game theory is a branch of applied mathematics… that… attempts to mathematically capture behavior in strategic situations , in which an individual's success in making choices depends on the choices of others. Famous: Nash equilibrium
  • 418 01 context for ai

    1. 1. Artificial Intelligence <ul><li>Context for AI </li></ul><ul><li>Dr. Gene Rohrbaugh </li></ul><ul><li>Fall 2008 </li></ul>
    2. 2. Artificial Intelligence <ul><li>A field devoted to building intelligent agents </li></ul>
    3. 3. Textbook Questions <ul><li>“Surely computers cannot be intelligent-they can only do what their programmers tell them.” Is the latter statement true and does it entail the former? </li></ul><ul><li>“Surely animals cannot be intelligent-they can only do what their genes tell them.” Is the latter statement true and does it entail the former? </li></ul>
    4. 4. Characteristics of Intelligence <ul><li>Ability to Communicate </li></ul><ul><li>Creativity </li></ul><ul><li>Internal Knowledge </li></ul><ul><li>Ability to Learn </li></ul><ul><li>World Knowledge </li></ul><ul><li>Goal-Directed Behavior </li></ul><ul><li>Self Awareness </li></ul>
    5. 5. Intelligence <ul><li>Hard to evaluate! </li></ul><ul><li>!= consciousness </li></ul><ul><li>Sentience = conscious perception </li></ul>
    6. 6. Turing’s Test
    7. 7. Foundations <ul><li>Philosophy </li></ul><ul><li>Mathematics </li></ul><ul><li>Psychology </li></ul><ul><li>Computer engineering </li></ul><ul><li>Linguistics </li></ul>
    8. 8. Philosophy <ul><li>Socrates-Plato-Aristotle </li></ul><ul><li>Dualism (Descartes) </li></ul><ul><li>Materialism (Leibniz) </li></ul><ul><li>Empiricism (Bacon, Locke) </li></ul>
    9. 9. Mathematics <ul><li>Syntax of formal logic (Boole, Frege) </li></ul><ul><li>Semantics of formal logic (Tarski) </li></ul><ul><li>Undecidable </li></ul><ul><li>Intractable </li></ul>
    10. 10. Undecidable <ul><li>“In any language expressive enough to describe properties of natural numbers there are true statements whose truth cannot be established by any algorithm.” </li></ul>
    11. 11. Intractable <ul><li>“A class of problems is called intractable if the time required to solve instances of of the class grows at least exponentially with the size of the instances.” </li></ul>
    12. 12. Linguistics <ul><li>Panini (350 BC): one of the earliest recorded analytical descriptions of a language </li></ul><ul><li>Plato (in Cratylus): relationship between language and reality -- conventionalism & naturalism </li></ul><ul><li>Chomsky -- formal properties of languages </li></ul>
    13. 13. AI Fields <ul><li>Agents </li></ul><ul><li>Search methods </li></ul><ul><li>Game theory </li></ul><ul><li>Knowledge representation </li></ul><ul><li>Communication </li></ul>
    14. 14. Searching <ul><li>Degrees of intelligence </li></ul><ul><li>Size of KB </li></ul><ul><li>Use of heuristics </li></ul>
    15. 15. Games <ul><li>Simple models are useful for exploring complex problems </li></ul><ul><li>Checkers, Chess, Go </li></ul>
    16. 16. Communication <ul><li>With human users </li></ul><ul><li>With other agents </li></ul><ul><li>With electronic sources/DB’s </li></ul>
    17. 17. What sorts of issues will we consider? Religious How does this impact our understanding of ourselves? Ethical Who is responsible for their behavior? How do we make sure they behave? Philosophical Are they alive? Technical How do we build intelligent agents?
    18. 18. AI Applications <ul><li>Expert Systems: provide human-like expertise </li></ul><ul><li>Robotics: mimic physical human processes </li></ul><ul><li>Natural-Language Processing: use human languages </li></ul><ul><li>Intelligent Tutorials: facilitate human learning </li></ul><ul><li>Computer Vision: mimic human sensory(visual) process </li></ul><ul><li>Virtual Reality: mimic human reality inside a computer </li></ul><ul><li>Game Playing: beat humans in games, e.g. chess </li></ul><ul><li>Augmented reality: modify percepts for human user </li></ul><ul><li>Intelligence Amplification: enhance human cognition </li></ul>
    19. 19. AI Techniques <ul><li>Fuzzy Logic: Facilitate solutions to human vagueness problems </li></ul><ul><li>Genetic Algorithms: Use data to find “optimal” solutions </li></ul><ul><li>Neural Networks: Use data to predict outputs or interpret inputs </li></ul>
    20. 20. AI in the news… <ul><li>Http://www. Aaai . Org/aitopics/pmwiki/pmwiki . Php/aitopics/ainews </li></ul><ul><li>Sandia equips robot for bomb disposal </li></ul><ul><li>Experimental robots `learn' right and wrong </li></ul><ul><li>Trainers school automaton like a child </li></ul><ul><li>Freestyle skiing robot may be useful in space </li></ul><ul><li>Nerve cells are linked with electronics </li></ul><ul><li>The once and future nanomachine </li></ul><ul><li>Biology outmatches futurists' most elaborate fantasies for molecular robots </li></ul><ul><li>Robo-gator to protect fish farms from thieving birds </li></ul>
    21. 21. Future AI Technologies <ul><li>Automatic Programming </li></ul><ul><li>Evolutionary Programming </li></ul><ul><li>Biological Artificial Neural Networks </li></ul><ul><li>Real Time Planning and Re-Planning Systems </li></ul><ul><li>Intelligent “learning” Agents </li></ul><ul><li>Autonomous robots, infobots, and nanobots </li></ul><ul><li>Biometric Security Systems </li></ul>