The Technology Behind Cognitive Computing

What does “Cognitive Computing” really mean? You can infer that it has something to do with a computer that can reason. But is it like the androids and robots we have all seen in movies, or is it more Silicon Valley hype? Well, neither. A few multi-billion dollar companies are trying to bring this technology to us, but it is not quite what you may expect. My goal is to demystify this topic and others in a series of articles covering advanced technology concepts like the “Cognitive Era”. In each post, we will explain the progress technology is making, how it works, and what it means, all while providing practical examples of how it can be used for your business, your life, and your family.

Cognitive Computing Industry

Jumping right in, the Cognitive Computing industry has been very active as of late, with IBM being one of the leaders in this industry. Starting in 2014, IBM began betting heavily on these technologies, including a massive $3B fund for research and development in Synthetic Brains and Quantum Computing. While this was occurring, the IBM Watson group was formally created, and they are spending more than $1B a year integrating the Cognitive Era into their company. Our company, Neon Mobile, has partnered with IBM Watson to deliver cognitive computing technology as an integral part of our Customer Engagement Platform solutions for businesses of any size and scale.

What Is Cognitive Computing?

But what exactly is Cognitive Computing? From an excellent IBM Watson White Paper “Computing, cognition and the future of knowing”, cognitive computing is defined as, “systems that learn at scale, reason with purpose and interact with humans naturally. Rather than being explicitly programmed, they learn and reason from their interactions with us and from their experiences with their environment.” Well, while that is easy to write, let’s hope that understanding what that meant is easier than interfacing with R2D2 (it will be, don’t worry).

Let’s look at the definition more closely. “Learn at Scale” – scale translates to millions of lines of code and exabytes of information that has to be accessed in tenths of seconds, but this code isn’t traditional code; it does not add numbers or paint computer screens. Instead it analyzes, and references, information like a graph with nodes similar to the structure of the brain in a way that gives you answers to questions or problems. It is not just held to yes or no answers, but now it has the ability to give a probabilistic answer based on what it knows, for example it may be 85% certain.

Now on to a more interesting word, “Learn”, which is the secret sauce for the cognitive technology companies. This is allowing the ability to assimilate large amounts of information to create synthesized neural networks that map the information and its relationships. These relationships can be viewed as branches on a tree where each branch has many paths, their thickness represents the weight or probability that the branch is the correct one, and based on that thickness, the branch holds more weight when making a decision. Creating these trees of information, in essence growing the tree and pruning the appropriate branches, is still one of the hardest problems to tackle in cognitive computing as it requires massive amounts of computing power. As the abilities of computers advance to handle new sources, including real-time visual, digital, and emotional information, it will yield far superior interaction capabilities in the near future.

Next, to “reason with purpose” – reasoning is something that we are unaccustomed to from a computer standpoint because having a computer understand us as humans with intentions is just now an emerging capability. A great example of this ability is the technology in our mobile phone assistants, Siri and Cortana. But how do you reason with a purpose? You do this by integrating data about yourself: your movement, your actions, your behavior, which will in turn create a purposeful response. By learning your habits, the responses can be personalized; you enjoy spicy food, you live in a certain area and don’t stray out of a certain radius often, you stay up late at night, you have high blood pressure, your stress level is high, the list can continue indefinitely. But by combining all of these data points of some new and previously inaccessible information, having the ability to reason now becomes possible. A cognitive based engine can recommend what food you should eat, how to become healthier, and help you reduce your stress.

Last but not least, cognitive computers are able to “Interact with humans naturally”. The keyword is “naturally” – naturally means varying the context of the conversation fluidly so it doesn’t feel like you’re talking to a machine. The days of yelling at telephone dialogues telling you to “Press 1 for frustration, or Press 2 to yell” is over. Cognitive computing is allowing people to have conversations within boundaries of a topic, like businesses or sports, without any hassle. There is not a HAL-9000 yet, but there will be eventually! Conversational systems that use technology similar to Siri and Cortana will start appearing everywhere. At Neon, we have partnered with IBM Watson to create our own conversational engine that any business can use with minimal setup. The future of natural conversation between computers is upon us, and will be everywhere very soon.

In the next articles, I will discuss how you can use this technology in your business and life, taking a deep dive into some of the more interesting capabilities of today’s emerging technology that will help you compete in the coming decade.

Leave a Comment

Contact Us

We're not around right now. But you can send us an email and we'll get back to you, asap.