Machine Learning vs Artificial Intelligence

At our recent GDS CMO Data Insight Summit in Sarasota, where marketing insiders were talking shop about data analysis and the latest cool gadgets, buzz words like Machine Learning (ML) and Artificial Intelligence (AI) were tossed about. But wait!! Amazon Echo is considered Machine Learning? Not Artificial Intelligence? Aren’t the two concepts interchangeable? Turns out they’re not and I needed to get my tech talk straight before using them in a conversation at the next cocktail party.

The Basics:

According to Wikipedia,  Artificial Intelligence is usually defined as the science of making computers do things that require intelligence when done by humans. The simplest example of AI? The self-driving car, a reality already in the making.

Machine learning, again from Wikipedia, is the process of getting computers to act without being explicitly programmed. An example of machine learning is when I make a typo while searching Google and it says, “did you mean……?” What was of most interest here to our summit’s marketing leaders is we can make machines learn things. Right now machines are learning a lot about us and marketers are paying close attention.

Robert Wellborn, AVP data scientist at USAA and one of the presenters at our summit says the most succinct way to understand the two is this: machine learning is a specific branch of AI and needs to be understood for its nuances. Induction based reasoning and memory based reasoning are all forms of AI but ML is about testing, adjusting the test statistics, seeing the result and testing again.

Wellborn says, “Machine Learning is not a new concept and was actually first referenced in Alan Turing’s writings in 1948.

The influx of data being obtained through our internet use and our smart phones has exploded in the last five years. Marketers are learning how to use all that data to tailor the customer experience.

“It’s petabytes of data coming in at one time, which is an amazing stream” says Wellborn. The property and casualty insurance company is looking at data from its customers in order to customise.

But first, what the heck is a petabyte? And how much information is that? According to LifeWire.com 1 petabyte is equivalent to over 4,000 digital photos per day over your entire life! 1 petabyte is equal to over 1 quadrillion bytes. So yeah, marketers are getting a crazy amount of information about customers every single day.

“The big change in the last five years is the amount of data that we have” says Rex Briggs, CEO Marketing Evolution. “We’ve gone from a few thousand variables per consumer to hundreds of thousands of variables. It all interacts with the type of statistical approaches you take and machine learning, artificial intelligence and the automation of pushing those insights into action has also been a major change facilitated by cloud computing.”

Artificial intelligence and machine learning are making their mark in the world of marketing. And although different, they’re both giving data miners extensive insights into consumer behavior.

“We are now able to use that data to provide meaningful content for our members when they need it,” explains Wellborn, “and not waste anyone’s time. It’s become a priority for us.”

Data is now driving the customer experience. Marketers are just trying to figure out how to use petabytes of information, break it down and tailor it to you. Oh and by the way, feel free to throw petabytes into your next cocktail conversation, I know I will.