What are intelligent systems?

For many healthcare professionals, the topic of artificial intelligence may seem daunting at first.  I know for a fact that artificial intelligence isn’t taught during those Methods and Logic in Medicine lectures that are commonly slept through by medical students.  As a result, the words “artificial intelligence” conjure up images of robots firing laser beams from their fingers, or self-driving cars that can sense their surroundings, or machines that can interact with other machines to play soccer.  How did these machines get that way, and where does one begin to try and understand artificial intelligence and intelligent systems?

In my first post, I am going to try and alleviate the fear and uncertainty around the words “artificial intelligence.”  If we peel back the layers and start just with some simple definitions, we can gain a new framework for approaching the topics of artificial intelligence and machine learning.

Let’s start with some statements about artificial intelligence and intelligent systems:

  • Artificial intelligence is a branch of computer science.  In 2013, the world-renown Association for Computing Machinery (ACM) and the Institute of Electrical and Electronics Engineers (IEEE) jointly published university curriculum guidelines for the computer science field (available here).  The identify 19 sub-fields within computer science, which include things such as Information Management (a.k.a. Databases), Discrete Mathematics, Software Engineering, and Information Security.  One of the subfields is called “Intelligent Systems.”
  • Artificial intelligence can be defined as “the study of solutions for problems that are difficult or impractical to solve with traditional methods,” according to the joint ACM/IEEE committee in their Intelligent Systems description.  This is a broad definition, but it is easy to see how it makes sense.  Problems that are challenging such as speech recognition, text recognition, language translation, object identification, and event prediction cannot be solved with traditional algorithms or brute force, but instead require special types of algorithms that go beyond traditional number crunching.
  • They go on to elaborate: “It is used pervasively…in the design and analysis of autonomous agents that perceive their environment and interact rationally with the environment.”  This is where the ACM/IEEE definition meshes with that used in movies.
  • The committee further goes on to subdivide “Intelligent Systems” into “sub-sub-divisions”, such as Basic Machine Learning, Advanced Machine Learning, Robotics, Perception and Computer Vision, etc.  Each of the “sub-sub-fields” can usually be taken as separate courses in college, or on Coursera (for us late-bloomers).

These points are not to underplay the complexity that lies behind modern intelligent applications.  Popular commercial apps such as Amazon’s Alexa and IBM’s Watson are built on millions of man-hours of work.  Nevertheless, approaching artificial intelligence and intelligent systems using a framework helps to break down the field if you are interested in learning more about it.

So, in conclusion, if you were intimidated by the words “artificial intelligence” as I was when I first heard the words used, be intimidated no more.  In later blog posts, I plan to explore the components of artificial intelligence in more detail.

Leave a Reply

Your email address will not be published. Required fields are marked *