Welcome to part 1 of our new series on "The Building Blocks of AI". In this five part series, we will aim to deconstruct the component pieces that put the intelligence in Artificial Intelligence. Interested in learning how they work, what they can do for your business, and how to evaluate potential technology partners? Read on.

Tweet about this on TwitterShare on FacebookShare on LinkedInShare on Google+

There is a lot of hype around artificial intelligence (AI) , and frankly, you should believe it! Gartner predicts AI will power 85% of customer service interactions by 2020, and has the potential to transform nearly every facet of modern business. Organizations across countless industries—from incumbents to startups—are eager to get their hands on some of this rocket fuel. The blocker? AI feels intimidating in the same way The Cloud did in the early 2000s… It’s an abstract idea, not a technology you can integrate or a product you can buy. So where do you start?

First, let’s unpack the term itself: AI is actually many different technologies working in different combinations to develop insights and act on them.. automatically. It’s only when you package those technologies into a product, with a specific purpose, that you can start to realize the real world benefits of this “great disruptor.” Let’s meet the component technologies that go into an AI system and learn how they can help your business today. First up: NLP.

Natural language processing (NLP) is the ability of a computer program to understand human speech as it is spoken or written, which is an incredibly complicated task. A machine can be programmed to recognize vocabulary and even irregularities such as idioms, but understanding sentiment—what we actually mean when we talk—is much more difficult (imagine trying to determine a stranger’s mood by reading a few text messages). That ability has to be learned.

Why is this important?

If the machines can’t translate what we say into a language they can understand (data), then they can’t do a whole lot with it. If the program doesn’t recognize the difference between, “This product is so great!” and “This product is soooo great,” it can’t qualify that response as positive or negative, and it certainly can’t respond with any accuracy.

How NLP fits into the larger AI picture

NLP is an essential ingredient in technologies like opinion mining, and it’s what enables cutting edge chatbots to effectively triage and respond to customer service requests. But it also powers basic things like intelligent search, content archiving, and help menus. It also works in reverse: NLP is essential to automating the generation of content too… if you play fantasy sports, you might be interested to know that much of the editorial content on those sites was bot-authored.

What can NLP do for you?

NLP can help you turn massive amounts of passive customer data (emails, recorded calls, message boards) into real and actionable insights. If you’re evaluating technologies that use NLP, make sure they are not just big dictionaries: they need to be able to learn and evolve their understanding to meet the unique needs of your business and your customers. How? Enter our next technology: Machine Learning.

Check back next week for part 2 in our series on deconstructing AI, and learn more about our AI-powered chatbot offerings here!

Posted by on Friday, April 7, 2017.


Leave a Reply