COMP316A: Artificial Intelligence

1
COMP316A Artificial Intelligence

• Lecturers
• Bernhard Pfahringer bernhard_at_cs.waikato.ac.nz
• Eibe Frank eibe_at_cs.waikato.ac.nz
• Email subject should include COMP316 in
  subject heading, or it is likely to be eaten by a
  spam filter

2
AI work at Waikato CS

• Machine learning/ data mining group
• Geoff Holmes, Eibe Frank, Bernhard Pfahringer,
  Mark Hall, John Cleary, Ian Witten, Tony Smith,
  Mike Mayo
• Associations with Reel 2 (John Cleary)
• Natural language processing Tony Smith
• Robotics Margaret Jefferies
• Intelligent tutoring systems Mike Mayo

3
Overview and History of AI

• History http//www.stottlerhenke.com/ai_general/
  history.htm
• Sample definitions
• The automation of activities that we associate
  with human thinking, activities such as
  decision-making, problem solving, learning
  (Bellman, 1978)
• the study of mental faculties through the use of
  computational models (Charniak McDermott,
  1985)
• the study of how to make computers do things at
  which, at the moment, people are better (Rich
  Knight, 1991)
• AIis concerned with intelligent behavior in
  artifacts (Nilsson, 1998)

4
Historical Attempts - Talos
(by 850 BC, when described in the Iliad) Talos, a
strong man, created by the god of smiths,
Hephaestus, whose job was to protect Crete by
casting stones at passing ships, thus warding off
pirates.
5
The Golem

• GOLEM literally means "inert matter" or
  something shapeless. In the Hebrew myth of the
  golem, a lump of clay was brought to life by
  appeal to the name of God. One of the major golem
  legends is the one attributed to Rabbi Judah Loew
  of Prague (1513-1609). Loew created Golem to
  protect him and his people. He did so for a
  while, but, as in other versions of the golem
  legend, the artificial clay man ran amok, and
  Loew was forced to remove the name of God from
  his mouth, and with that, Golems life.

6
Frankenstein

• Novel by Mary Shelley, 1818, describing attempt
  by scientist Victor Frankenstein to create life.

7
The Turk
8
RUR

• 1921 play, R.U.R (Rossums Universal Robots) by
  Czech author Karel Capek
• "CHEAP LABOR. ROSSUM'S ROBOTS." "ROBOTS FOR THE
  TROPICS.  150 DOLLARS EACH.""EVERYONE SHOULD BUY
  HIS OWN ROBOT." "DO YOU WANT TO CHEAPEN YOUR
  OUTPUT?  ORDER ROSSUM'S ROBOTS" 
• In the days when Czechoslovakia was a feudal
  society, "robota" referred to the two or three
  days of the week that peasants were obliged to
  leave their own fields to work without
  remuneration on the lands of noblemen. For a long
  time after the feudal system had passed away,
  robota continued to be used to describe work that
  one wasn't exactly doing voluntarily or for fun,
  while today's younger Czechs and Slovaks tend to
  use robota to refer to work thats boring or
  uninteresting. (E. Rich)

9
Acting humanly the Turing Test

• Presented by Turing (1950), Computing machinery
  and intelligence
• Can machines think? -- Can machines behave
  intelligently?

10
Turing Test

• 1990 Loebner Prize established. Grand Prize of
  100,000 and a Gold Medal for the first computer
  whose responses are indistinguishable from a
  human.
• (Solid 18 carat, not gold-plated like the Olympic
  "Gold" medals)

11
How much computing power does it take?
From Hans Moravec, Robot Mere Machine to
Transcendent Mind 1998.
12
How much computing power exists?
From Hans Moravec, Robot Mere Machine to
Transcendent Mind 1998.
13
Brief history of AI

• 1943 McCulloch Pitts, boolean circuit model
  of the brain
• 1949 Hebbian learning model (neural networks)
• 1950 Turings Computing Machinery and
  Intelligence
• 1951 first neural network computer (Minsky
  Edmonds)
• 1952 Samuals first checkers-playing program
  (on TV, 1956)

14
1950s - 1966 great expectations

• 1956 Dartmouth workshop term AI adopted
• 1950s marked by beginning of AI Hype
• 1950 Turing predicted that in about fifty years
  "an average interrogator will not have more than
  a 70 percent chance of making the right
  identification after five minutes of
  questioning".
• Newell and Simon predicted that "Within ten years
  a computer will be the world's chess champion,
  unless the rules bar it from competition."
• 1950s, 60s automated language translation was
  just around the corner

15
great expectations

• Why was the AI hype overly optimistic? Failure
  to understand
• Need for knowledge (about domains of application,
  common sense)
• Scalability and problems of complexity and
  exponential growth
• The need to perceive the world, rather than
  working in simulation or limited contexts
  (microworlds)

16
great expectations

• 1958 McCarthy, Lisp programming language
• Start of hype that AI is language-related if
  its in Lisp or PROLOG, its AI
• Later (1980s) special purpose Lisp or Prolog
  machines

17
1966 - 1973 Dose of reality

• 1969 book, Perceptrons (Minsky Papert)
  proves that perceptrons incapable of representing
  (and therefore learning) over many classes of
  problems funding for neural nets disappears
• 1969 Bryson Ho, multilayer NNs discovered,
  but languish until 1980s
• Natural language processing the limitations
  sink in
• the spirit is willing, but the flesh is weak
• Becomes, the vodka is good but the meat is
  rotten

18
1969 - 1979 knowledge based systems

• Weak methods use general purpose search,
  reasoning to string together steps to find a
  solution
• Not scalable too many possibilities at any step
• Strong methods include domain-specific
  knowledge to help focus

19
1979 - 1990 Glory times

• Expert systems rule!
• Rule based, human-readable systems that serve as
  expert practitioners in narrow tasks (ex
  MYCIN, diagnosis of blood infections)
• ES for supplemental feed for range cows
• Rule 1 if protein is low and energy is low then
  feed is high
• Rule 2 if protein is low and energy is high then
  feed is low
• Rule 3 if protein is high and energy is low then
  feed is high
• Rule 4 if protein is high and energy is high then
  feed is low

20
Glory times

• 1986 return of the neural networks
• Parallel Distributed Processing, McClellan
  Rumelhart
• A very clever precedent a readable book
  freely available software

21
Glory times

• 1981 - 1991 Japan announces Fifth Generation
  project to build intelligent computers running
  Prolog
• US, European funding agencies pour money into AI
  so as not to be left behind
• Major hype in software now with AI!
• AI firms pop up everywhere

22
1991 - 1995 AI Winter

• Hype meets reality Fifth Generation fizzles out
• Much angst will we ever have jobs again?
• Re-focusing of AI in industry goal is no longer
  to do AI, but to apply AI techniques when
  appropriate to problems
• Re-focusing of goal moving from trying to
  produce software to replace a human (a la expert
  systems, automated translation), to trying to
  support a human (for example, machine learning as
  support in extracting information from data)

23
1995 - present Agents rule!

• Agents embedded in real environments with
  continuous sensory inputs
• WWW a major impetus to development
• AI becomes (more of) a science emphasis in the
  research literature on empirical experiments,
  statistically analyzed development of common
  testbeds for comparing alternative algorithms

24
A cynical summary

• What is AI? AI is whatever doesnt work
• Computer vision -- image processing
• Automatic programming -- compilers
• A less cynical summary AI is an attempt to
  understand and support people and high-level tasks