AI and the Future

There are many names that describe the field of computing which seeks to give computers the ability to deal with the real world intelligently.  Among them are artificial intelligence, machine learning, or just plain intelligence.  Computer scientists are now working on methods to build computers that can operate on a level past basic logical reasoning, like mathematical operations or games of chess.  Artificial intelligence instead seeks to build machines that can learn from the environment and real people, as well as learn from the mistakes it makes along the way (very much like real human beings have done for all of our existence).  In fact, a number of computer systems already achieve this kind of real-world interaction and application. Anti-lock brake systems, airplane autopilot systems, and even email spam folders that differentiate legitimate mail from spam are all examples of this type of technology.

So how do artificial intelligence machines work? It all starts with what is called a model, or an initial prediction for how the machine should work, and it is given by a human being to the computer.  For example, a teacher wanting to use artificial intelligence to determine the proper amount of time to study for an example will first enter a model like 2 hours for a C, 3 hours for a B, and 5 hours for a C.  From there, real data is input to the computer, which is then used by the “learner” to change the model accordingly for a more accurate prediction.  This simple cycle is repeated over and over again, with minute adjustments being made by the computer along the way to achieve incremental change.  Through observing more and more real data, the machine will develop the best model it can.

Google wants to make Star Trek technology a reality.

One application of machine learning is image recognition.  Obviously, computers do not have eyes and thus cannot identify pictures the way that we do.  As a result, machine learning can step in to give computers the ability to differentiate a picture of a cat from a hot air balloon.  To do this, a human must first enter a model for the characteristics that might make up a picture of a cat as opposed to the hot air balloon.  After that, the computer is given test sets of pictures so that it can develop a better sense of what makes up specific pictures, and change the model accordingly. 

Right now, many companies are devoting its resources to develop artificial intelligence machines that will function as natural human assistants.  Google, for example, likens its Google Assistant project to the intelligent computer famous from Star Trek.  “If the Assistant works as well as Google hopes, all you might have to do is say, ‘O.K., Google, I need to go to Hong Kong next week. Take care of it,’” according to a New York Times articial on the project. This technology has the power to make sci fi movie technology a reality.

Pictures & References:

http://marketingland.com/how-machine-learning-works-150366

http://www.nytimes.com/2016/09/29/technology/google-assistant.html