What is Artificial Intelligence (AI)?

According to Dictionary.com, artificial intelligence is “the capacity of a computer to perform operations analogous to learning and decision making in humans…”.  Some prevalent examples are the ability to understand someone’s voice (natural language processing), learn from previous mistakes, errors, or patterns (machine learning) and recognize objects (machine vision).

Natural language processing (NLP) allows computers to interpret what a user is saying and the information the user is trying to get from the software (CRM, Inventory system, database, etc.).  While this doesn’t seem too difficult, being able to understand different accents, dialects and jargon can be very difficult for a computer.

One example that doesn’t include variations in audio, like accents, is that there are ten different ways that a customer can ask your parts department if they have a radiator: Do you have a radiator? I need a radiator. Do you have a radiator for my 2007 Ford? Do you have a radiator in stock? etc.. Computer programs are written to eliminate these type of errors by giving the operator choices (drop down boxes, checkboxes, etc.), but artificial intelligence allows us to get this information through natural dialog.  This is the technology behind Siri, Alexa, Cortana and Google Home.

With machine learning, we can turn large amounts of data into meaningful insights by understanding patterns to predict future outcomes. Additionally, when an algorithm knows the outcome, it can reference that against the original prediction to get better for the next prediction, hence the name machine learning.  In returning to our example of ordering a radiator, machine learning allows the program to better understand what a user is saying through repetition.  So, as more people interact with the system using different dialects, accents, and questions, the technology (Sophi) gets better and better at understanding what it is the customer is trying to accomplish to provide a better user experience.

Lastly, machine vision, allows computers to recognize objects by analyzing the pixels in an image.  We can easily see how this technology allows autonomous vehicles to recognize a lane, other vehicles, and humans.