If a new technology doesn’t live up to its hype, companies can waste a lot of money investing in it. Analysts at Gartner understand this, which is why they developed their famous Hype Cycle. As they note, “Gartner’s Hype Cycle is a graphical depiction of a common pattern that arises with each new technology or other innovation. … The five phases in the Hype Cycle are Technology Trigger, Peak of Inflated Expectations, Trough of Disillusionment, Slope of Enlightenment and Plateau of Productivity.”
So where does machine learning (ML) fall on the hype cycle? At the very least, it’s on the slope of enlightenment and rapidly climbing towards the plateau of productivity. As Eric Siegel (@predictanalytic), a former computer science professor at Columbia University, writes, “Machine learning … is genuinely powerful and everyone oughta be excited about it.”
Beyond the Hype
Bratin Saha, Vice President and General Manager at Amazon, agrees that machine learning has advanced beyond the initial hype surrounding it. He observes, “Machine learning has been around for decades, but for much of that time, businesses were only deploying a few models and those required tedious, painstaking work done by PhDs and machine learning experts. Over the past couple of years, machine learning has grown significantly thanks to the advent of widely available, standardized, cloud-based machine learning platforms.” As a result of this growth, he notes, “Today, companies across every industry are deploying millions of machine learning models across multiple lines of business.” He continues:
“The next phase of machine learning will deliver what scientists could only dream of: industrializing and democratizing machine learning. With purpose-built machine learning platforms and tools that can systematize and automate deploying machine learning models at scale, we’re on the cusp of a major shift that will make it possible for all enterprises — not just the global Fortune 50 companies — to use this transformative technology and become truly disruptive. … To deliver on the vision and promise of decades of work, machine learning models need to solve complex business problems, provide actionable insights in real time, and become integrated into operational systems and processes. … The good news is that machine learning is industrializing and moving away from the hype to become a mature engineering discipline formed along two vectors: purpose-built machine learning platforms and specialized machine learning tools.”
Saha goes on to note that machine learning depends on the availability of quality data and understanding how models make their predictions. Many machine learning models are “black box” systems whose predictions are unexplainable. At Enterra Solutions®, we leverage the Representational Learning Machine™ (RLM) created by Massive Dynamics™. The RLM can help determine what type of analysis is best-suited for the data involved in a high-dimensional environment and it is also a “glass box” system. Obtain the right data, identify a specific problem, and apply the right analytics model and you get the results for which you are looking. From work with our clients, we know machine learning can generate significant benefits. Saha, looking at a number of industries, agrees machine learning can produce significant results. He concludes, “Machine learning industrialization is paying real dividends.”
Machine Learning Use Cases
The staff at Matillion note, “The applications and uses of machine learning are vast and diverse — and they’re all around us, every day.” And journalist Angela Scott-Briggs insists companies can leverage machine learning “to skyrocket your business’s growth.” Below are some of the ways machine learning is currently being used in the business world:
1. Automation. Scott-Briggs notes, “One of the most renowned and well-documented uses of AI is in automating daily tasks. The tasks that are automated don’t necessarily have to be mundane either, ML can even automate routine tasks like data recovery, auditing, reporting, monitoring, and much more.”
2. Search Engine Results. The Matillion staff explains, “Every time you type a search term into Google, machine learning algorithms analyze your behavior to refine the future delivery of results. For instance, if you spend a significant length of time on a site that wasn’t highly ranked on the initial results page, the Google algorithm will likely bump that page higher for similar or related searches in the future.”
3. Recommendations. Streaming services and e-commerce have made recommendation engines extremely valuable. The staff at CIO Applications notes, “Deep learning models enable ‘people also liked’ and ‘just for you’ recommendations [in] entertainment, retail, job search, travel and news services.”
4. Fraud Detection. The Matillion staff notes, “Using machine learning models, banks and other financial institutions can identify transactions that fall outside typical parameters — such as purchase amount and user location — and alert you when unusual activity occurs.” The CIO Applications staff adds, “Machine learning regression and classification models have overthrown rules-based fraud detection systems that have a high number of false positives when flagging stolen credit card utilization.”
5. Gradual Optimization. According to Scott-Briggs, “Just like every good thing, machine learning takes its own time to show any tangible improvements. The optimizations done by ML are gradual in nature, and their impacts start showing up as the AI learns more and more. You’ll have to look deep and find areas of your company where there are large data sets that can benefit the most from ML.”
6. Risk Predictions. Scott-Briggs writes, “ML can take into consideration many more variables than a human manager can, and then it can compare the data with past trends and future predictions to understand what kind of risks the company might have to face.”
7. Chatbots and Digital Assistants. The Matillion staff notes, “When you chat with an AI-based assistant to resolve an issue online, a trained machine learning model is at work, providing appropriate responses based on your input.” And, the CIO Applications staff adds, “Digital assistants are empowered by natural language processing, a machine learning application that allows computers to process text as well as voice data and ‘understand’ human language.”
8. Customer Targeting and Online Advertising. The CIO Applications staff explains, “Machine learning … models are capable of evaluating the content of a web page — the topic and nuances such as the author’s attitude or opinion — and serve up advertisements customized to the visitor’s interests.” Scott-Briggs adds, “Targeting your customers specifically while shooting an advertisement towards them is one of the most effective strategies that you can use. This is the reason why the advertisements you see are based on your recent search history and interests. If you utilize machine learning to shape your customer’s experience as per their interests, then you can rest assured that you’ll witness tremendous growth.”
9. Spam Filters. “By analyzing characteristics in subject lines,” the Matillion staff notes, “body content and return addresses, machine learning algorithms help protect your in-box from unwanted emails.”
10. Customer Research. Scott-Briggs writes, “Researching what your customers need and want is one of the most crucial aspects of running a successful business. Traditionally, managers rely on market research to understand the demands of their market, eventually filing in the holes that exist in the demand-supply chain. … Researching is a crucial aspect of running a business, and at no stage should you stop trying to detect the trends in the market if you want your company to stay profitable.” The Matillion staff adds, “Customer retention service providers rely on machine learning models to identify customers who may be ready to take their business elsewhere. If you’ve stopped using a credit card and suddenly received an email offer for a lower APR, your credit card provider is likely attempting to boost customer retention with the help of a machine learning-based platform.”
11. Gauging the Market. Markets change all the time. Scott-Briggs observes, “To detect a new and upcoming trend, one must be able to analyze a huge dataset. … This is where machine learning shines the brightest, as it can easily analyze large quantities of data within a short period of time. The trends that can be detected aren’t very limited either, you can view the trends in everything that uses and leaves some sort of data trails like sales, hiring, and even advertising.”
12. Specialized Machine Learning Cases. The Matillion staff notes that machine learning can be used for specialized use cases like: candidate screening; real estate evaluation; education applications; and medical image processing. The point is: Machine learning is now mature enough to be applied to many business processes that can enhance the bottom line.
Once machine learning reaches the Plateau of Productivity, new ways of putting machine learning to work will undoubtedly emerge. Most pundits agree, machine learning has progressed well beyond the hype. Siegel concludes, “I gotta say, it wins the Awesomest Technology Ever award, forging advancements that make ya go, ‘Hooha!’.”
 Staff, “Hype Cycle,” Gartner Glossary.
 Eric Siegel, “Why A.I. is a big fat lie,” Big Think, 23 January 2019.
 Bratin Saha, “Machine learning is moving beyond the hype,” InfoWorld, 14 September 2021.
 Staff, “The Uses of Machine Learning and the Benefits for Your Enterprise,” Matillion Blog, 12 October 2021.
 Angela Scott-Briggs, “Machine Learning Strategies That’ll Significantly Increase Business Efficiency,” TechBullion, 2 November 2021.
 Staff, “Use Cases of Machine Learning,” CIO Applications, 31 May 2021.