Skip to Content
July 19, 2018 Article 4 min read
You hear it everywhere: AI is the future. What you may not know is that it’s already here, you probably use it daily, and it’s poised to make your job a whole lot easier.

Picture of a computer storage/server room.

For years, we’ve been hearing that artificial intelligence (AI) is the future. If you’ve just been skimming headlines, all that conversation might seem like little more than nebulous ideas about AI’s risk to American jobs, self-driving cars, and the eventuality of a robot uprising. What you may not know is that AI has several practical, nonthreatening, everyday uses for businesses and individuals alike — and you’re probably already interacting with it on a regular basis.

If you’ve ordered a Lyft or Uber, sent a text that your phone auto-corrected, enabled an email spam filter, watched a show recommended by Netflix, deposited a check through your banking app, or used a virtual assistant like Alexa or Google Home, you’ve interacted with AI.

If you operate a business, don’t underestimate its usefulness. AI might be the future, but it’s also the present, so forget the idea that there’s plenty of time to figure it out. AI can offer a major competitive advantage, and if you fail to invest in new technology, you might fall behind.

Before we explain how to use AI in your business, let’s start with the fundamentals.

What is AI?

Artificial intelligence, an umbrella term for all machine intelligence, describes a machine’s ability to perform human (typically cognitive) functions. This means that the machines have the ability to learn, reason, and/or problem solve. With advancements in sensor technology, machines now have the ability to perceive their environments and interact with their surroundings, most famously in the form of autonomous vehicles and virtual assistants.

In a business context, AI is great for analyzing data in four primary ways:

  • Descriptive analytics tells us what happened (and doesn’t always require AI).
  • Diagnostic analytics focuses on uncovering why events occurred from hidden patterns in data.
  • Predictive analytics tells us what’s most likely to happen based on past data.
  • Prescriptive analytics pairs descriptive and predictive to recommend action based on what might happen.

Prescriptive analytics has the greatest potential and is often considered the “future of big data.” Over the last couple years, companies have shifted focus from predictive to prescriptive, and many are already putting it to use. For example, we helped one hospital to develop a predictive algorithm that identifies high-risk patients that’s now being enhanced through prescriptive analytics to recommend specific courses of action based on each patient’s unique clinical history and health needs.

How do machines learn?

Machine learning is a big area of focus that falls under the AI umbrella, and there are three major categories:

Supervised learning works exactly how it sounds — the machine is trained from selected data samples, which may be generated with human involvement. We’d generally consider the machine “taught” once the algorithm is generating predictable results with unfamiliar data sets.

Law firms are using supervised learning to quickly summarize contract terms; lenders are using it to predict default rates on loans. The risk with supervised learning, however, is bias — the machine can only operate as well as the data that taught it.

Unsupervised learning uses algorithms designed to discover patterns without training data. Retailers use unsupervised learning to uncover hidden patterns about the buying patterns of their customers; for example, customers who buy tortilla chips are also very likely to buy salsa. There’s no need to teach the algorithm with a training dataset — it discovers the patterns automatically.

Reinforcement learning uses algorithms that continue to learn over time as they’re exposed to more data sets. Some people compare it to mastering a video game where you collect more data each time you play until you reach the end. The automotive industry is using reinforcement learning — and many other AI techniques — in autonomous vehicles.

Who will benefit most from AI?

In a business environment, the most accessible application is data analytics. Data analytics benefits anybody who wants to put their data to work — and you need less data than you may think — so most industries have a lot to gain.

AI benefits anybody who wants to put their data to work — and you need less data than you may think.

For retail buyers, this could mean analyzing several seasons of point-of-sale data to predict the best product mix for the upcoming year. For manufacturers, it might involve a program that identifies flaws in parts to improve quality controls. And for legal firms, it might include teaching the machine to comb through a mountain of case documents to find a single paragraph.

In other words, AI can potentially help you create better work, in less time, using fewer resources.

AI can help you create better work, in less time, using fewer resources.

What now?

Believe it or not, you don’t have to be a tech whiz to put AI to use. It takes a data scientist to create the algorithms, but software companies are producing software that makes it easy for the average person to produce something valuable. Get an idea of what tools might work for you by searching “AI software for [your industry here].”

Despite the ease of use and availability of new AI options, you need to make sure you protect your data, especially if you plan to use it to analyze client information. Do yourself a favor and consult an expert — when making an investment like this, you could lose a lot of time, money, and possibly data if you don’t get it right. An analytics consultant can help you get started with developing an analytics strategy for using AI in your business and developing your first AI algorithm while an information technology consultant can help you find the right product to suit your needs.