Artificial intelligence is used to develop computer systems that demonstrate characteristics of intelligent behavior. Expert systems make it possible for a novice to perform at the level of an expert in very specific situations. Learn more about both of these systems in this lesson.
What Is Intelligence?
Perhaps a few years ago you watched on TV when two champions of the popular game show Jeopardy! were defeated by an artificial intelligence computer system called Watson. The computer was able to process human speech, search huge databases very quickly and reply in a human voice. The company IBM took about four years to build Watson. They used 90 computer servers, 360 computer chips and very sophisticated software.
While Watson took home the $1 million prize, more important is the fact that the underlying technology can be used in the real world. For example, in 2013, the Watson software system was implemented in hospitals to help diagnose lung cancer and to recommend treatment. The computer certainly does not replace the medical staff but can assist them in being more accurate and consistent in their jobs.
Watson is one example of artificial intelligence. Artificial intelligence, or AI, systems include the people, hardware, software, data and knowledge used to develop computer systems that demonstrate characteristics of intelligent behavior. Intelligent behavior has a number of different components to it, including the ability to:
- Learn from experience and apply the knowledge gained from those experiences
- Handle complex situations
- Solve problems when important information is missing
- Determine what is important
- React quickly and correctly to a new situation
- Understand visual images
- Process and manipulate symbols
- Be creative and imaginative
These are all things that we as humans are naturally good at since that is how our brains work. On the other hand, a basic computer system is good at performing simple tasks very quickly, such as doing calculations with numbers or searching through large amounts of data. So, how do we create a computer system with intelligent behavior? Let's look at some of the different forms of AI to see how this is accomplished.
Types of AI
Artificial intelligence includes a wide range of systems that can replicate human decision making for certain types of well-defined problems. In robotics, mechanical or computer devices perform tasks that require a high degree of precision or are very tedious or hazardous to humans. The controlling software in robotics uses AI. Robotics is used in many applications, from assembly lines for cars to the robotic arm on the International Space Station.
One critical aspect of AI is to interact with humans. This is where technologies such as natural language processing come in. Natural language processing allows a computer to understand and react to statements and commands in a human language, such as English. Many automated telephone services now include the option to speak your instructions instead of selecting options on your keypad. Some types of mobile phones and car navigation systems also allow you to speak your commands.
Learning systems are major elements of many AI systems. Learning systems use a combination of hardware and software to allow a computer to change how it functions or reacts based on feedback. Some computer-based games have this built in. If a computer does not win, it remembers not to make the same mistake twice. This is one reason why it is so hard to beat an advanced level computer chess game. It has learned from a large number of known chess matches.
An increasingly important aspect of AI systems is the use of neural networks. A neural network is a computer system that tries to simulate some of the functionality of the human brain. Using a mesh-like structure somewhat similar to a brain, neural networks can process many pieces of data simultaneously and learn to recognize patterns.
An alternative to trying to developing computer systems with true intelligent behavior is to use an expert system. Expert systems make it possible for a novice to perform at the level of an expert in very specific situations. Expert systems take a much more limited view of what intelligence is and use a set of detailed rules. These rules are based on the documented expertise of one or more individuals. An expert system simulates the reasoning and decision making of these experts.
Software is used to collect and store the experiences and knowledge of human experts from various professional fields. This is referred to as the knowledge base. This knowledge base must be developed for every specific application. For example, a system to assist with the evaluation of financial statements would be very different from a system used to run diagnostics on the electronic components of a vehicle.
In addition to a knowledge base, an expert system uses an inference engine. This component looks for information and relationships in the knowledge base to provide answers, predictions and suggestions. This is similar to the reasoning of a human expert when being presented with a particular problem.
An expert system also needs a user interface so a novice can actually use the knowledge base and inference engine. Quite often, the user is presented with a series of questions. For example, a junior accountant is asked to enter a number of details about a financial statement using a series of questions, and the expert system presents a number of inconsistencies or points of concern that warrant further attention.
Developing an expert system typically requires the collaboration of several different people. Domain experts are the people with the expertise the expert system is trying to capture. Knowledge engineers are the technical specialists with training in the design and development of expert systems. Knowledge users represent the individuals or group who will be using the expert system. This collaboration ultimately makes it possible for knowledge users to make decisions similar to domain experts.
Let's look at an example of an expert system. Consider a bank that issues loans to small businesses. A senior loan manager is able to decide relatively quickly on whether to approve a particular loan application based on the business plan, the financial history of the applicant and other factors. An expert system makes it possible for a junior loan manager to use this expertise. A computer-based user interface asks a series of questions about the loan application and walks the junior loan manager through the analysis of the loan. The analysis is based on the domain expertise of more senior staff. This particular expert system is very specific for small business loans. For example, an expert system to evaluate mortgage loan applications would need a different knowledge base.
- Artificial intelligence (AI) systems include the people, hardware, software, data and knowledge used to develop computer systems that demonstrate characteristics of intelligent behavior.
- Artificial intelligence includes a wide range of systems that can replicate human decision making for certain types of well-defined problems.
- Expert systems make it possible for a novice to perform at the level of an expert in very specific situations.
After you have finished with this lesson, you'll be able to:
- Identify the components of intelligent behavior
- Describe how artificial intelligence systems replicate intelligent behavior
- Explain what an expert system is and describe the components that go into one