Please rotate your device.

Our website uses cookies to ensure you get the best experience while you’re here.


Breaking Down Artificial Intelligence (AI)

By: SEI Team

Steampunk style professional robotic vacuum cleaner machine with hand brushes on orange wall background

With all the noise and hype around artificial intelligence (AI) advancements, business savvy individuals should be aware of the opportunities this technology presents to any organization. While there are many naysayers of AI’s impact on our world, the list of ever-growing enhancements that many thought would never come to fruition are a result of true technological advancement. In this blog, we will breakdown the real benefits around this technology starting with a basic definition of AI, take a look at two of the more business-ready AI fields, and provide suggestions of how your company can use them to support your AI journey.

Complexities of AI

AI is a complex field that can be confusing due to its breadth. On one side of the spectrum you have sci-fi-like machines that “perceive, reason, and act” – think Hollywood Human Level AI’s (HLAI), such as The Terminator or Wall-E. On the other side, you have specifically focused machines, or AI agents, that “perform functions that require intelligence when performed by people.” It is on the latter side of the spectrum where wealth of practical applications of AI exist for businesses to leverage. Typically, in this application, after an AI solution has been found for a problem, over time the problem ceases to be considered AI simply because of its newfound commonplace functionality (e.g., speech recognition, email spam filters, supply chain route planning). More recent developments that have not had enough time to transition into commonplace status include the Roomba robotic vacuum cleaners and Predictive Data Analytics. Looking farther down the road, there are exciting problems that AI will solve, creating real and dramatic impact on our world.

To break down the world of AI, we will focus on two fields that are applicable to medium and large enterprise businesses – Cognitive AI and Machine Learning. Both of these fields have progressed over the last few years from theoretical experiments in the academia and startup domains into mature areas with strong capabilities and tools.

Cognitive AI

Cognitive AI systems exhibit human-like intelligence using the processes of learning, reasoning, and memory. These processes are interrelated – reasoning works to understand and make decisions, which are determined to be right and wrong through learning and then stored in memory for future reasoning. Cognitive AI itself can be further divided into many subcategories (e.g., Design and Creativity, Planning, Common Sense Reasoning, and Analogical Reasoning), all of which are then further subdivided into ever more specialized areas. A rule of thumb to understanding the scope of Cognitive AI is that it attempts to mirror rational human thinking. Just as we would break a problem down into smaller pieces or solve a problem based on past experiences, so too does Cognitive AI. IBM’s Watson is a Cognitive AI tool that leverages many of these subcategories, such as natural language processing, knowledge representation and reasoning to provide solutions that range from creating chat bots to helping find individualized treatment options for cancer patients.

Machine Learning

Machine Learning focuses on learning through data. Machine Learning can be used on its own to create models and analyze data or combined with other fields, such as helping Cognitive AI learn. Machine Learning can be further subdivided into Supervised, Unsupervised, and Reinforcement Learning and again sub-divided from there. Supervised Machine Learning occurs when you teach the computer using data with answers, such as providing a list of wines with varying characteristics and ratings into an algorithm and creating a model to predict the rating of a previously unknown bottle of wine. Unsupervised learning occurs without the answers. This can include analyzing large amounts of data, such as clustering related customer data by transforming it, simplifying it, pulling out key information, or making the data richer for additional machine learning tasks. Reinforcement Learning is where an agent learns how to behave based on rewards. An example would be a robot that explores its environment and is rewarded when performing a specific action. Examples of Machine Learning include things like Netflix movie recommendations, credit card default predictions, stock trading algorithms, robots that walk, and more.

Approaching AI

When deciding where a company should focus its AI efforts, they should first develop a basic understanding of the core capabilities that AI provides and then align to the AI fields that provide the most benefit. To prioritize and test initiatives for implementation, use the following approach:

Strategy 1: Develop a deeper understanding of AI

  • Debunk the hype, understand what AI can, and can’t do, and what’s realistic for your company
  • Create an AI team to champion AI opportunities, aligned with Innovation groups where they exist
  • Think through goals and scenarios that can help your business grow and how they might be achieved with AI

Strategy 2: Identify likely areas that would benefit such as the following examples:

  • Data Analytics is a clear and readily available opportunity to take action on your data by predicting which customers are most likely to purchase to which options and products to sell
  • Robotic Processes Automation provides the ability to automate repetitive rule based tasks providing significant cost savings, increased quality, and increased efficiency for work such as processing invoices and many other back and middle office related activities
  • Cognitive AI, such as Watson, can be used for a variety of human understanding scenarios such as monitoring a facilities repair needs, diagnosing patients, assessing claims, and preparing content

Strategy 3: Develop a plan, prioritize, and begin to implement AI solutions

  • Many of the above examples can be rolled out in phases, start with quick win, high value targets
  • Evaluate vendors and in-house talent and tools and include in planning expanding organizational capabilities where needed
  • As the capability within the company is grown, shift the AI focus from one of experimentation to that of a standard, given function within the business

The key is to start now and experiment while building capabilities and delivering value. AI isn’t something in the far-flung future, it’s here now. Whether or not this is the final tipping point to an AI immersive future, businesses should fully embrace the capability’s readily available benefits to stay ahead of the pack that is undoubtedly attempting to do the same.