In a recent Health Imaging article, two medical thoughts leaders predicted that the patient safety movement will increasingly advocate the use of algorithms over humans. Massive data sets along with advances in technology are on the way to making profound improvements in healthcare, as clinical pathways change and medical errors go down. In the field of radiology, for example, computer-aided detection for screening mammography can replace the second opinion of a radiologist, according to the New England Journal of Medicine. Every industry, and everyday life, will be impacted by machine learning. However, algorithms and data are not all equal when it comes to how the modern person works and lives life. In the world of law, it is easy to see how machine learning systems can be trained to review case history and help draft legal briefs. But can machines lay down the law? Try to imagine this: a self-driving car get into an accident and the case goes to court that is decided by a machine. In cases where binary is being prosecuted, litigation should become faster and more accurate through judicial automation and digital decision making. But is it possible – or even allowable – for a machine to judge a medical system for poor computer vision and improper interpretation of digitized images?
The question that comes out of these examples is: under what conditions should algorithms and data replace people? Computerized Numerical Control (CNC) machines have been around for decades where software essentially drives drills, lathes, cutters and other tools in combination with automation such as robotics for manufacturing. Digitization has had a profound impact on the factory floor and jobs have been lost to automation over the past 30 years, close to 6 million according to the U.S. Bureau of Labor Statistics. But manufacturing output has been on the rise, and for the United States, it has more than doubled over the past three decades according to the National Association of Manufacturers. In fact, Deloitte Global and the U.S. Council on Competitiveness predicts that the U.S. will become the most competitive manufacturing nation in the world in 2020. But should a machine replace decision making when human inconsistencies and inclinations such as bias are baked into the software? Humans are fallible, and machines built to replace certain jobs may not be neutral. That may be fine for manufacturing but certainly not in a court of law, where justice is supposed to be blind.
Science fiction is now business fact, according to McKinsey and Company. They analyzed 2,000-plus work activities for more than 800 occupations and the results showed that currently demonstrated technologies could automate 45 percent of the activities people are paid to perform. It is not surprising that predictable work is on the top of the list and helps to explain why manufacturing has always been a poster child for digitization. On the other side of the spectrum is applying expertise to decision making, planning and creative tasks. Attorneys and appellate judges need not worry about robots replacing them any time soon, but legal costs could go down in the Digital Era. For predictable legal work, LegalZoom, the online technology company, is making starting a business, getting a divorce or protecting a family’s estate easier and lower in cost. Machine learning has been used to predict legal outcomes, something attorneys get paid for. A statistical model created by researchers at Michigan State University and South Texas College of Law was able to predict the outcome of almost 71 percent of U.S. Supreme Court cases. These kinds of machine learning applications can have tremendous social, economic and political ramifications. That’s why it may be best to have a human in the loop when using algorithm-based predictions.
All things going digital may also explain why there is no shortage of jobs in America today. As industries are being transformed by innovative technologies, skilled workers are needed to ease labor shortages in the digital economy. A Pew Research study of employment data shows that the job categories with the highest growth tend to require higher social skills, analytic savvy and technical prowess. An amazing 85 percent of American workers say having a detailed understanding of how to use computer technology is extremely important or very important. The key to success, according to Pew, is having higher levels of analytical skills, including computer use. One does not have to look further than the cybersecurity area to fully appreciate the need for skilled people. Software-based hacks and attacks frequently make the front page news. The most recent Arbor Networks Worldwide Infrastructure Security Report (WISR) shows a dramatic growth in volumetric DDoS attacks and how the Internet of Things is changing the threat landscape. The number of jobs in information security is going to grow tenfold by 2025 and illustrates a massive human resource need in the future, considering there were 209,000 cybersecurity jobs unfilled in the U.S. in 2015. And cybersecurity job growth is 12 times faster than for all other jobs, according to Burning Glass Technologies in Boston, which uses big data analytics to help close the skills gap between job supply and demand.
Of course, the IT infrastructure is what the digital economy depends on, 24/7. It is the common denominator for many aspects of work and social life. Accordingly, managing digital transformation is strategic for business and IT leaders and service performance and availability are the most important transformational metrics. While the winds of change continuously swirl around every enterprise, the IT organization must discover now (find ways to take advantage of digital), plan for the future and make the journey. And accomplish all that within the context of reducing business risk and enabling innovation with confidence.
The point that needs to be made here is the criticality of service assurance in a digital world and the need for proactive service triage as automation and algorithms become more and more a part of our lives. Boston Consulting Group has predicted that by 2025 as much as a quarter of jobs currently available will be replaced by either smart software or robots. If digital disruptions happen, then businesses and their customers will suffer. The costs are real including tarnished brand and lost revenue. According to an IDC survey of Fortune 1000 companies in 2015, the average cost of application failure per hour is $500,000 to $1 million and a Forrester Consulting report said IT resolutions cost $11 million per year per company. The IT infrastructure is not getting simpler but rather growing in complexity due to virtualization, cloud, unified communications, security, micro-services, and more. In our interdependent, digital economy where speed and agility reign, it is not unheard of to have DevOps people deliver and deploy new software thousands of times a day.
The bottom line for the IT organization is visibility into the service performance and impact on business. They can’t afford not to have unrestricted visibility when things go wrong, especially considering the hundreds of millions and even billions of transactions a day that digital enterprises need to support. What becomes abundantly clear for lines of business or departments is winning and keeping customers happy depends on accelerating digital innovation and assuring automation works, 24/7. Thus, business leadership must have a high level of confidence in people, processes and service assurance platform.
The opportunity for things breaking grows with the inherent IT complexity tied to digital transformation. Ask any IT professional and they will point to a long list of service-impacting problems that manifests themselves as one way phone calls, unable to login, failed transactions, slow applications, can’t save or retrieve files in the cloud, garbled video, systems not working, sensors opening up new cyber-attack vectors, and so much more. If services fail, then the time to innovate grows, and as a result slows down progress. As such, the IT infrastructure and people that manage it are not only tasked with assuring service delivery, but successful outcomes. Whether we like it or not, our quality of life depends on these guardians of the connected world.
But how can they reduce risk? In a world where there is no off, real-time information and intelligence is strategically important to getting ahead of problems before they impact how we work, live and play. Leo “the Lip” Durocher, the swashbuckling major league baseball manager, understood the importance of visibility when he said; “I never questioned the integrity of an umpire. Their eyesight, yes.” Blind spots in the IT infrastructure are simply bad for us. By harnessing intelligence from traffic data and other sources, the IT organization is able to assure the automation we take for granted today in banking, transportation, retail, manufacturing, utilities and more. But real-time information and intelligence is also the basis for accelerating digital transformation and assuring tomorrow’s algorithms deliver value above and beyond what people can do today.