Essential AI: A brief introduction to AI terms and applications

For a long time, artificial intelligence was something spoken about in terms of the future. The impact it could make or will make someday. So it might be surprising to a casual observer just how much has changed in recent years. Because with AI today, “will” has become “is.”

In one recent survey of 260 global businesses, 80% of companies are already building or using AI systems. As they say, the future has arrived.

With AI taking an increasingly central role at businesses of all sizes, you may already be using AI in one form or another. If not, it’s only a matter of time until you are. Once you’re accustomed to using AI tools, you’ll likely find that your job is a lot more dynamic and fulfilling than it was before. One of the core benefits of AI is taking over low-level tasks such as scheduling and data entry, allowing people to focus on more creative and rewarding activities.

In the same way that professionals needed to come up to speed on software and desktop computers in the 1990’s, today we all need to start learning more about the key terms and concepts of AI. With that in mind, we put together this overview to get you started.

Artificial intelligence. Before we get into too many specifics, let’s start with something we’re all very familiar with: software. Software is a very broad term that encompasses many different categories. There’s productivity software like word processors and spreadsheets, which are similar to but distinctly different from database software or CAD/CAM software.

As with the term software, artificial intelligence is a catch-all that can refer to a diverse set of uses and techniques, with some of the big ones being neural networks, machine learning, and deep learning. We’ll dive into those in just a moment.

The common thread running through each of these technologies—and the trait that makes AI distinct from traditional software—is that AI systems can interpret, analyze, and learn from data in ways similar to human cognition. AI doesn’t function exactly like a human brain, but to the outside observer that’s what appears to be happening. Here’s a quick way to picture the difference. With traditional software, everything a user might do with an application needs to be programmed line by line to ensure the intended results. Advanced AI systems, in contrast, are built around algorithmic models with the power to generate results that aren’t necessarily known in advance. An AI system can produce surprises; traditional software generally doesn’t. It’s this flexibility that puts the “intelligence” in artificial intelligence.

With that in mind, let’s look at a few important AI techniques that illustrate various approaches to building advanced AI systems.

Machine learning. Broadly speaking, machine learning is a computer science field that gives machines the ability to analyze a set of data, draw conclusions about the data, and then make predictions when presented with new data—without being explicitly programmed to do so. Learning is the key term here. The more information a machine learning system receives, the more accurate its predictions become. Machine learning has been used to power computer vision, natural language, and data analysis applications, and it’s increasingly being explored as a tool to aid professionals in a wide range of industries.

Medical doctors, for instance, can use machine learning programs to more quickly and accurately diagnose diseases. The machine analyzes many thousands of images of both healthy and unhealthy organs, and learns to pinpoint disease indicators by contrasting the two categories.

Neural networks. The concept of neural networks dates back to 1944, with its popularity ebbing and flowing in the ensuing decades as theory and potential wrestled with the limitations of available computer hardware. Neural networks are so named because their structure is inspired by the neural connections of the human brain and their intelligence comes from the ability to analyze countless data inputs to discover context and meaning. Those inputs pass through a layer of algorithms that “learn” as more information is passed through. As a shorthand, you can think of neural networks as a specific technique for doing machine learning.

You may not realize it, but you encounter neural networks on a regular basis. Many image recognition programs, for instance, run on neural networks. Algorithms at different layers of the program interpret details of a photo such as pixel brightness, lines and shapes, textures, and distinctive features. Working in concert, the individual elements combine to identify people and objects in photos.

Deep learning. Another approach within artificial intelligence that’s getting a lot of attention lately is deep learning. Deep learning is an advanced machine learning technique that uses multiple neural networks and massive amounts of computing power. Modern computers  are exponentially more powerful than those for previous decades, allowing for “deeper” layers of neural networks and therefore faster, more dynamic applications.

Deep learning systems are at the heart of emerging technologies as diverse as real-time translation and autonomous vehicles.

Artificial intelligence solutions in everyday life

Once you understand the basics of AI functions, you begin to see artificial intelligence everywhere. AI is reshaping the world around us, driving progress, helping businesses glean more valuable insights from their data, and ultimately empowering them to make smarter decisions.

Here are a few places where AI is already well at work:

Product recommendations

We’re all familiar with product recommendation features, whether you’re visiting an online retailer or a streaming video service. You browse or purchase one item, only to see a list of related products appear somewhere on the page. Perhaps you even begin seeing ads and receiving promotional emails showcasing similar items.

Some of those recommendation engines are powered by deep learning systems that parse data on users’ browsing and buying behavior to anticipate other products they might be interested in. One interesting application is the way Netflix offers dynamically generated movie thumbnail images to entice you to watch new films and TV series.

Don’t be surprised if you start having increasingly personalized content experiences as well. Marketers are embracing AI as a path to achieving better engagement with potential customers. Rather than pouring over analytics to determine which types of content should be served to which audience segments, AI-enabled smart content automatically adjusts messaging based on the audiences’ interaction history. AI-backed content can also self-adjust to become more easily discoverable, and smart chatbots can recommend content to website or app visitors based on existing data.

Biometric authentication

Imagine logging into your bank account not with a password or fingerprint but with an iris scan or vocal verification test. Given the value of financial data, it’s unsurprising that banks are investing heavily in biometric authentication systems.

AI tools like computer vision and natural language processing can significantly enhance security. It’s a lot harder to fake someone’s vocal pitch than it is to hack their password, and it’s harder still to mimic their eye’s unique patterns. Many such systems that use facial recognition to verify customers via selfies are already in use. They’ll only become more sophisticated and secure as AI systems are able to analyze people’s faces, voices, eyes, and even their behavioral patterns simultaneously to determine if people are who they say they are.


AI is shaping up to be a formidable tool for fighting cyberattacks, which represent a growing threat in areas such as healthcare, finance, and insurance. Hackers who steal people’s personal data can use it to access their financial accounts and commit insurance fraud, wreaking havoc in victims’ lives. AI systems learn to spot anomalies in databases and transactions, and they can run 24/7, so there’s no downtime when hackers can easily break in.

Given the enormous amount of data generated every day, human efforts must be augmented by AI to truly defend against cyber criminals. This is an increasingly relevant use of the technology, particularly as more and more companies store consumer data digitally and conduct financial transactions online.

Although artificial intelligence has existed as an idea for more than seven decades, we’re only beginning to see its power and potential. The tools built on technologies such as machine and deep learning will transform workplaces and even how we think about our careers and lives. With advanced AI on their side, the businesses of the future can be more productive, efficient, and secure.