AY Honors/Artificial Intelligence/Answer Key

1. What is artificial intelligence?

Artificial intelligence is man made intelligence using machines and software designed to perform given task and mimic the thinking process of a human being.

“The science of making machines do things that would require intelligence if done by men” – Dr. Marvin Minsky (MIT)

2. On your own or with a group, develop a chart board that outlines a brief history of artificial intelligence. Prepare and give an oral presentation on your activity.

The focus of this chart board is for the pathfinder to begin their research into the history of artificial intelligence and its path of acceptance as an academic discipline.

Should include more than modern AI

• 4th Century BC – Aristotle invents syllogistic logic and deductive reasoning
• 1206 – Al-Jazari designs what is beleved to be the first programmable humanoid robot
• 1642 – Pascal creates first mechanical calculating machine
• 1662 – Sir Samuel Morland devises Arithmetical machines
• 1673 – Leibniz improves Pascal’s machines to add division and multiplication and devises Calculus to determine how reasoning can be decided mechanically
• 1854 – George Boole develops binary algebra representing the laws of thought and creating Boolean expressions
• 1936 – Alan Turing proposes the universal Turing Machine
• 1943 – McCullock and Pitts publish “A logical Calculus of the Ideas Immanent in Nervous Activity” laying the foundations of neural networks
• 1956 – John McCarthy coined the term “Artificial Intelligence”
• 1957 Newel & Simon (Carnegie Mellon) demonstrate the General Problem Solver
• 1958 – John McCarthy invents LISP
• 1964 – Danny Bobrow defends his dissertation at MIT on Natural Language Processing
• 1968 Minsky and Papert publish “Perceptrons” – demonstrating limits of simple neural nets
• 1969 Shank (Yale) defined conceptual dependency model of natural language understanding
• 1972 – Comerauer develops PROLOG
• 1976 – Lenat’s dissertation with the Automated Mathematician program demonstrates computer learning and discovery of interesting conjectures
• 1980 – First AAAI (American Association of Artificial Intelligence) held at Stanford
• 1987 – Minsky publishes “The Society of Mind” a description of the mind as a collection of cooperating agents
• 1997 – Deep Blue defeats Garry Kasparov in a chess match

3. What is the ultimate goal of artificial intelligence research?

The ultimate goal of artificial intelligence is to create a device that is capable of making independent external decisions using a series of self driven internal instructions. In other words, a part of its objective is to make a computer perform more like a human.

4. What is an android?

An android is an intelligent machine with a human appearance. Continued advancements in robotic technology have lighted the way for more improved designs that better impersonate human beings.

5. With regard to the field of artificial intelligence, be able to define the following terms.

a. General Intelligence

a machine that is capable of displaying human-like general intelligence such as self awareness and consciousness.

b. Social Intelligence

a machine that has the ability to express social skills and emotional skills based on the reactions of other people that it may come into contact with.

c. Creativity

The processes of a machine gathering data that it knows and hypothesizing independent/spontaneous new information that is relevant to the problem

d. Learning

a process of a computer evaluating input and extracting new knowledge or heuristics from the observation of existing data.

e. Motion

a machine's ability to move and locate objects, including itself.

f. Planning

the ability of a machine to set goals, and then move forward to achieve those goals.

g. Perception

a machine's ability to use its own inputs like sensors, cameras, and microphones to develop its own conclusions.

h. Heuristic

a machine's ability to search and locate the shortest path.

i. Pattern Recognition

a machine's ability to make identification based on a series of inputs.

j. Neural Network

used for machine learning, giving a machine the ability to ‘think’, based on selected inputs.

k. Natural Language Processing

The ability for a machine to process spoken or written text and then devise and understand the context of the information being passed. This is more than easy key word searches, but is actually centered on contextual information being garnered from text or speech.

l. Knowledge Engineering

The process of interviewing experts in a given field and extracting the rule base/heuristics which allow them to process information and make conclusions. This rule base is then coded in the AI system to give the computer the captured expertise of the human.

6. Give three real world examples of how artificial intelligence is used to help society.

Artificial Intelligence is used in a variety of industries and fields.

In the manufacturing industry, the use of robots has increasingly become a standard practice. Selected functions are given to various robots, each working with the other as part of a larger goal. For example, AI can be used to maintain the cleanliness of a bottling factory. One robot is used to scan every bottle for its strength. Once a weak structured bottle is detected, it signals to another machine to knock that bottle off of the line. After a period of time, broken glass has accumulated in the designated broken glass area. Another robot waits for a preprogrammed amount of glass to accumulate, when it then comes out to clean all of the broken glass. This is all done without human intervention.

In the finance industry, billions of dollars flow through the financial institutions. Due to off shore accounts, money laundering, and the vast number of ways employed to defraud the banks, AI is used to flag transactions that will require human intervention. These transactions may have errors that were created by the sender that will trigger an alert.

Any company that is customer service driven will also find use to employ AI. AI can be used as a first level support alternative to answer some of the most basic of questions. This approach allows the more difficult questions and problems to be addressed by human service representatives.

An enormous amount of information is natural language processed by computers each day to evaluate the context of the conversations and/or writing to assess potential threats posed to the country.

7. What are some of the limitations to artificial intelligence? Be able to explain at least three.

The inability to learn from other people by either accepting or rejecting what they say as facts.

The inability to understand when to use proper means of communication at a given instant; weather it be seeing, reading, writing, or speaking.

As of this writing, the overall speed is slow. This is due to the vast amount of code involved to complete a simple task. For example, a programs designed to play a game can be very large due to the number of conditions that may exist in a given moment, i.e., where to move next in a checker game.

Machines are not sentient, they do not create spontaneous thought, cannot find humor or sadness in situations and cannot distinguish the importance of information on a personal level

8. What are some basic human abilities that artificial intelligence can not exhibit?

Artificial Intelligence is its various forms are prone to activities that they can not exhibit. For example, AI robot can not:

• Dream, let alone understand or explain what it dreamed
• Acknowledge the presence of the almighty God
• Express emotion based on its surroundings
• Develop spontaneous thought devoid of conditional data
• Understand sarcasm, irony or humor
• Find the bridges for unrelated data without human intervention

9. Give a basic definition of an expert system?

An expert system is a branch of artificial intelligence. It is used to solve very complex problems by emulating the decision making process of a human being. It utilizes knowledge engineering to encapsulate the rules of operations from human experts, than captures those rules into heuristics and rapidly processes mass amounts of data through the rule set. Forward chaining takes the data through the rule sets marking what conclusions can be drawn, Backward chaining takes the data through the rule set and if conclusions are drawn looks at the marked rules that were not fired to determine if the new conclusions will draw any further conclusions. Hence it is non-linear processing of data in the same manner that human processes information sets.

10. What are some of the advantages of an expert system?

An expert system:

• renders an unemotional response at all times
• reduces danger for humans
• can be designed to have expertise in many areas
• can explain in detail how a conclusion was derived
• can draw conclusions many times faster than a human

11. Give a real world example of how an expert system is used in society.

Expert systems are used in the gaming industry; particularly for games that offer human vs. computer options. When playing against a computer, the expert system (the actual decision making engine of the game) is called to determine the next move given the current circumstance. In a game of chess, the professional moves of a grand master chess champion would be programmed, thus forming the expert system.

12. On your own or with a group, discuss the importance of artificial intelligence and the role it plays in society. Prepare and give an oral presentation on your findings.

Pathfinders are encouraged to seek a better understanding of artificial intelligence by discovering how it is being applied in real world scenarios and its impact in society.