will guide the executive in selecting the right person and provide the Champion with skillsets and tools needed for implementing the first analytics project and scaling the Analytics Culture.
An executive's job is to manage risk, not avoid it. Yet many executives are too risk-averse and choose not to make decisions because the risk of failure blinds them to see the opportunity for success. While information is nearly always imperfect, employing AI and analytics gives vision to the future that mitigates risk for better decision-making. This book is for you, the executive and aspiring executive, to arm you with the knowledge to lead your organization to become an analytics powerhouse.
With this introduction, we welcome you to the Undiscovered Country—the future!
CHAPTER 1 A Primer on AI-Enabled Analytics for Business
Knowledge will forever govern ignorance; and a people who mean to be their own governors must arm themselves with the power which knowledge gives.
—James Madison1
Artificial intelligence (AI) dates back over 75 years. Alan Turing, a mathematician, explored the mathematical possibility of AI, suggesting that “humans use available information as well as reason in order to solve problems and make decisions,” and if this premise is true, then machines can do so too. This was the basis of his 1950 paper “Computing Machinery and Intelligence,” in which he discussed “how to build intelligent machines and how to test their intelligence.”2
So, what is artificial intelligence? Very broadly speaking, it is the ability of a machine to make decisions that are done by humans. But what does that mean, what does AI look like, and how will it change our lives and society?
We all know that AI, sooner or later, will be part of all businesses. But when it is part of the business is entirely dependent on what each executive knows and understands about AI and analytics. And here lies the chasm between the early adopters and the rest of the pack.
According to Grant Thornton's 21 May 2019 report “The Vital Role of the CFO in Digital Transformation,” the 2019 CFO Survey of Tech Adoption covered several technologies, including advanced analytics and machine learning. 38% of respondents indicated that they currently implemented advanced analytics, and 29% are planning implementations in the next 12 months. For machine learning technology, the survey results said that 29% had implemented it and 24% were planning to implement in the next 12 months. Impressive returns from the survey's sample set, and indicative of the priority of and accelerating trend in the adoption of analytics and AI throughout business. However, while conveying progress in its best light, this survey is a poor showing of a glass that is not even half full.
Implementations of AI are just scratching the surface, as projects have been highly targeted to only certain areas of the business and for certain tasks. So, while the movement to incorporate advanced analytics is in the right direction, there are many more failures than successes. This is disturbingly bad news, which we shall learn largely rests with executives. The good news is that AI and analytics failures are eminently avoidable.
Many executives lack clarity of vision and voice to how they will navigate their business, division, group, or department through the adoption of analytics and AI. Other executives think they know what AI enablement means but are often working from poorly defined terms or misconceptions about analytics. Their knee-jerk response is to hire consultants and buy AI-enabled analytics software without fully understanding how analytics will be used to drive decisions.
Cries of “We need better forecasting” and “What factors are driving our business?” and “We must get smarter about what we do” echo in boardrooms and executive conference rooms. But how exactly is this done? Not what, but how? The “what,” many an executive has read from a mountain of consulting reports; but the “how” is unclear and is why too many businesses are lagging in their adoption of AI and analytics.
In this chapter, we lay the foundation for this book by untangling terms and terminology with definitions and giving a ground-level introduction in select technologies (for the purpose of understanding, not to become experts). We will pursue a high-level discussion of AI, machine learning (ML), and analysis vs. analytics, followed by an explanation of business intelligence and data visualization and how these are different from analytics. We will introduce the application of AI-enabled analytics in the context of insights and the contrast between biased vs. unbiased predictions. Finally, we will position the importance of AI by discussing its ROI.
AI AND ML—SIMILAR BUT DIFFERENT
We see the widely used phrase “AI and ML” and conjure these as linked at the hip; but while related, they are not one and the same. First, AI is a superset, covering all that is considered artificial intelligence. The overarching concept of AI is simply a machine that can make a human decision. Any mode of achieving this human decision by a machine is thus AI, and machine learning is one such mode or subset of AI. Therefore, all ML is AI, but not all AI is ML.
Accordingly, ML is one form of AI. ML is a widely used method for implementing AI, and there are many tools, languages, and techniques available. ML engages algorithms (mathematical models) that computers use to perform a specific task without explicit instructions, often relying on patterns and inference, instead.
Another popular form of AI is neural networks that are highly advanced and based on mirroring the synapse structure of the brain. So, ML and neural networks are both subsets of AI, as depicted in Figure 1.1, as well as other forms of AI (that is, any other technology/technique that enables a machine to make a human decision).3
Figure 1.1 Superset and subsets of AI.
MACHINE LEARNING PRIMER
This section offers a brief orientation to ML. ML is a technique and technology that today requires specialized skills to use and deploy. ML is an AI engine often used with other tools to render the ML output useful for decisions. For example, suppose a bank wants to expand the number of loans without increasing the risk profile of its loan portfolio. ML can be used to make predictions regarding risk, and then the results are imported to spreadsheets to report those new additional loan applicants that can now be approved.
Large ML projects often involve the collaboration of data scientists, programmers, database administrators, and application developers (to render a deliverable outcome). Further, ML needs large volumes of high-quality data to “train” the ML model, and it is this data requirement that causes 8 of 10 ML and AI projects to stall.4 While ML is popular and powerful, it is not easy. Many new software applications are making ML use easier, but it is still mostly for data scientists.
Before an ML project can begin, its “object” must be defined: that is, what is to be solved. For example, suppose we want to predict which customers on our ecommerce website will proceed to check out (vs. those who exit before checking out). As presented in Figure
