CUTTER BUSINESS TECHNOLOGY JOURNAL  VOL. 32, NO. 7
  

Even if you have played no role in the design, development, or support of the applications that manage your per­sonal and professional lives, you know that software rules the world. Every aspect of your life is enabled by a suite of fixed or mobile software applications that more often than not live in the cloud. It’s safe to say that if you were separated or divorced from your apps, you would be unable to function. We can also define the rule of software by the integration of our personal and professional activities, which have strongly con­verged over the past decade in ways that often make it impossible to cleanly distinguish personal versus professional agendas.   

Let’s look at the extent of software’s rule today in five areas and where we expect it to be in five to 10 years. A “good/bad/ugly” lens will help us assess the trajectories and determine the role that software should — and should not — play in our lives.

The Five Themes

There are many ways to understand good, bad, and ugly software and what the reign of software will deliver in the next decade or so. This article covers five themes:

1. Process automation

2. Privacy and security

3. Enterprise software

4. Intelligent software engineering

5. Converged convenience

Theme 1: Process Automation  

Routine tasks — and even what appear to be the complex, deductive, inferential tasks that we associate with “knowledge” industries — will be automated by software bots of one kind or another; robotic process automation (RPA) will absolutely, positively eliminate jobs, careers, and whole professional existences. Indeed, it has been predicted that artificial intelligence (AI) (broadly defined) will eliminate 77 million jobs over the next 20 years: “By 2030, 75 million to 375 million workers (3 to 14 percent of the global workforce) will need to switch occupa­tional categories.”¹ Bloomberg has even developed a tool to help you determine if you’re likely to be automated. According to Bloomberg (based in part on research conducted at the University of Oxford), “Nearly half of all US jobs may be at risk in the coming decades, with lower-paid occupations among the most vulnerable.”² Compensation and bene­fits managers, auditors, accountants, credit analysts, loan officers, sales reps, truck drivers, administrative services managers, and even dental hygienists are at high risk and will most likely lose their jobs to automation. The same research suggests that (most) physicians, surgeons, (some) lawyers, financial managers, pharma­cists, teachers, and com­puter and information systems managers are among the professions least likely to be automated.³ The timing for all this varies. Some analysts believe significant pro­fessional displacement will easily occur by 2030, while others believe it will take longer — though not much longer.

Good, Bad, or Ugly?

If your job is in any of the at-risk categories noted above, you’re likely doomed. The important question is, “How long do you have?” This, of course, is the nagging question about all disruptive technologies and the impact they will have on the jobs market. A recent VICE News/HBO special, The Future of Work, presented demonstrations of disruptive technologies — such as self-driving trucks, expert legal systems, financial management tools, and surgical robotics — already hard at work.⁴

While there will be some lag due to regulatory and liability requirements (especially regarding autonomous vehicles), disruptive technologies are marching quickly toward deployment. Many industries consider all this very good. Professionals in the most vulnerable fields believe it’s bad. Some economists consider it all ugly since it displaces millions of professionals with no plan to relocate them into productive, well-paying careers. If ever there was an outcome dependent upon where you sit, process automation is it.

Theme 2: Privacy and Security   

There is no privacy. Everyone is under surveillance. Security is so weak that foreign governments are easily able to penetrate US elections. Software enables — and, to be fair, tries to combat — all these conditions. However, the trends here are anything but good.

Let’s start with security. According to the US Department of Homeland Security (DHS), threats are everywhere and growing. DHS believes that the US should “reduce threats from cybercriminals. In partnership with other law enforcement agencies, DHS must pre­vent cybercrime and disrupt criminals and criminal organizations who use cyberspace to carry out their illicit activities and leverage identified threat activity and trends to inform national risk management efforts.”⁵ The problem is enormous and growing faster than anyone can measure. Most computer scientists believe that no system is completely safe. Breaches are frequent — and frequently underreported.    

How about privacy and surveillance? If you’re on the grid, you’re under surveillance. If you tweet, blog, or post, you’re under surveillance. If you shop with credit cards, you’re under surveillance. If you rideshare, you’re under surveillance.

CCTV, smart TVs, Internet searches, social media, voice recognition/response systems, credit/debit cards, loyalty programs, facial recognition, image understanding, and even drones all enable surveillance. Within a few years, it will be possible for companies to profile nearly all of us from how we live our rich, full digital lives. As analytics improve, fewer and fewer digital indicators will be necessary to fully profile us. But surveillance will also empower government offices and agencies to profile individuals. Some of this will be good, such as enabling the pursuit, capture, and prosecution of criminals. But some of it will be ugly, such as what might happen when social, economic, and political enemies seek control, revenge, or worse. Make no mistake: the surveillance infrastructure is already in place and will only get wider, deeper, and stronger.

What’s next? Technologies such as AI, machine learning (ML), 5G, blockchain, crypto­currency, the Internet of Things, and wearables will make surveillance easier, faster, and complete. There’s no need to implant chips into our bodies, though some are doing so, because we’re immersed in digital trackers in our pockets, cars, homes, phones, TVs, appliances, thermostats, security systems, and, of course, our desktops, laptops, and tablets. We also know that leaving the digital grid is impossible. Surveillance is therefore inevitable.

Good, Bad, or Ugly?

This is an easy one — it’s not good, and, at times, is very ugly. The lack of privacy due to the rise of surveillance is bad and ugly. There are aspects that make sense, such as criminal and terrorist digital surveillance. But regardless of the percentage of good versus bad (or ugly) and the convenience software enables, software used to reduce privacy and increase surveillance definitely nets ugly.

Trajectory? Much uglier: total grid dependency and technologies such as facial recognition will finalize surveillance. Cybersecurity also looks bad and ugly. As more and more activities, processes, and assets move to the cloud, it will become increasingly difficult to secure transactions, especially since general awareness of the breadth, depth, and severity of threats is ill-defined and underappreciated, and because cyber­security funding, especially at the federal level, is incredibly inadequate. Both privacy and security are bad and headed toward ugly.

Theme 3: Enterprise Software

Who would undertake a five-year corporate software implementation project today? The failure rate for big “enterprise” software projects is downright scary. Depending on whose study you read, the enterprise resource planning (ERP) failure rate, for example, is anywhere between 50% and 75%. If even vaguely informed, executive management knows the project is likely to fail. Yet in the 1990s and early 21st century, there were still companies willing to try their hand with big software and prove they were not like the others who failed so spectacularly — until they too failed. Failure is the result of several trends and out­comes. One is control.

When a company embarks on a multiyear journey with an ERP or customer relationship management (CRM) vendor, they cede significant, if not total, process control to that vendor. ERP modules were originally designed to eliminate process chaos. Remember when “legacy” software was a barrier to scalability, not to mention how expensive it was to maintain? Moreover, a significant side effect of big software was the loss of process governance to the vendors that defined supply chain management, financial reporting, CRM, and other business processes for the companies they serviced.

The cloud also killed big software. Years ago, companies would implement huge enterprise software sys­tems in their own data centers. The early 21st century gave us the cloud, so the pain of forever implementation was avoided. But there are also smaller, cloud-based alternatives to big software that scale, integrate, and share process control through customization tools deliberately built into the modules. Small companies can find lots of incredibly inexpensive alternatives from vendors such as Zoho and Zendesk, among others. Many of these companies will grow, as will the incredibly inexpensive, cloud-based systems that scale and integrate right along with them.

Good, Bad, or Ugly?

The movement of huge enterprise software suites to the cloud is good. The adoption of smaller, cloud-based, microservices-based applications is also good. But “good” depends on where you sit: for obvious reasons, software consultancies preferred the endless on-premise implementation of huge enterprise software applications. Big software vendors were happier before the cloud offered alternatives that organizations could adopt relatively quickly. Note that the growing capa­bilities of business software applications are unquestionably good. Overall? Good. The consultancies and software vendors will adapt and rearchitect their huge software suites into smaller pieces. Watch how SAP, Oracle, IBM, and Microsoft, among others, adapt to the competition from smaller, “enterprise” vendors. In fact, many of them already have, even it if means selling smaller suites to their clients.

Theme 4: Intelligent Software Engineering 

We tend to think about functionality (i.e., what apps actually do) when we think about software. But where does software come from? How is it built? Will software help us develop software? Absolutely, and not just any old software: software will be designed and developed by intelligent — artificially intelligent — software. At the most basic level, smart software will automate many of the tedious steps in the software design and development process (e.g., testing). But the most significant impact will be felt in the auto-generation of code through the shadowing of human programmers and “learning” from their successes and failures — and then even deeper learning–based “programming.”

So what happens to programmers when all of this automation takes hold? The timing of intelligent software design and development is difficult to estimate, though it’s safe to say that within the decade much of this will be ready. Major software companies are investing heavily in intelligent software engineering, including SAP through its Leonardo Machine Learning Foundation.⁶ IBM, Oracle, and Microsoft, among others, are also spending heavily in the area. Programmers will evolve to support personnel. Application development lifecycles will be compressed. Programmers will become incredibly productive.

Good, Bad, or Ugly?

The software industry continues to grow. Intelligent, automated software design and development is good. In fact, artificially intelligent–supported software development will become one of the most promising application domains of AI and ML. Programmers will adapt over time and learn to exploit the support of intelligent software design/development assistants (which will eventually become leaders). One of the red flags is the ethics of automated software design and development — so-called ethical AI⁷ — and the value systems that enable expert and other intelligent systems to “decide” what to build. This is a larger issue for intelligent software engineering that could turn the assessment from good to bad, though probably not ugly.

Theme 5: Converged Convenience

We love streaming music and using location-based apps. We love tweeting and blogging. We love our project management tools and Microsoft Office 365. We love ridesharing. We also love Amazon and eBay. Love? Let’s just say that these and so many other apps are indispensable to our personal and professional lives.  

Lest we wax too poetic about the accessibility and functionality of these apps, remember that the greatest Trojan horse of the 21st century is the convenience our digital toys deliver: how easy it is to order anything we want from Amazon, how much fun it is to download music and books, and how effortlessly we can find a car, a house, and a date online. But what’s the tradeoff? Every time we avail ourselves of these conveniences, we reveal a little more about who we are and what we do, which is all stored and analyzed permanently for those who want to buy some insight into what we like, what we will buy, and how they should pitch to us. It all seems innocent enough until we assess what’s really happening. 

That said, software enables the integration and management of our personal and professional lives. Schedules, shopping, meetings, and grocery delivery can all be managed from single portals that manage multiple applications. Smart city applications help us navigate locations, and telecommuting applications help us work from home. There are countless others that keep our lives manageable and productive.

Good, Bad, or Ugly?

The problem with converged convenience software is that it’s really good, sometimes bad, and occasionally ugly. It’s also inevitable because convenience is undeniable. Trajectories show more of the same: our personal and professional lives will continue to converge, and our need for software that makes these lives easier will grow. Most users will sacrifice some — perhaps a great deal of — privacy and even security if you make their lives easier. Software is a huge and growing part of this “transaction.” Which presents a dilemma: Do we reduce convenience in exchange for privacy and security? Or do we sacrifice privacy and security for convenience? It’s likely that convenience wins for so many personal and professional reasons. The drivers are unstoppable as are the returns on personal and professional software investments. By 2030, the personal/professional convergence will be seamless and assumed.

Good, Bad, Ugly — or Something Else?

There can be no debate: software rules the world. The five themes discussed in this article suggest how software is rewarding, punishing, and threatening us — all at the same time. There are clear winners and losers in the reign of software. In process automation, com­panies win by reducing costs and increasing profit, but all while realigning and reducing whole professions. In enterprise software, some consultancies and big software vendors have reluctantly adjusted to microservice architectures and cloud delivery, and some have exploited both these features of newer enterprise software with inexpensive, scalable products and ser­vices. Intelligent software engineering will generate faster code as it changes the role of the traditional ­soft­ware engineer. Privacy and security are the losers in the reign of software. There’s too little awareness, focus, and funding, and it’s already way too late in the game. Security and privacy are bad, trending to ugly. Part of the explanation is traceable to our love of convenience and the software toys we refuse to divorce even though they compromise our digital freedoms. 

Above all else, we must acknowledge the inevitability of a world where software will continue to rule, and a world where the software kingdom will continue to expand. Some of this expansion will be good, some bad, and some ugly. Expansion also begins at earlier and earlier ages, with two- and three-year-old kids embarking into games and other digital toys on several mobile platforms. RPA seeks to automate as many corporate processes as possible. AI and ML will accelerate soft­ware development. Enterprise software will continue to shrink, spread, and scale. Privacy and security will yield to convenience. All of this is inevitable. The open question is, “What should we do about the kingdom of software?” Embrace it? Challenge it? Anything?

References

¹Vlastelica, Ryan. “Automation Could Impact 375 Million Jobs by 2030, New Study Suggests.” Market Watch, 4 December 2017.

²Whitehouse, Mark, and Mira Rojanasakul. “Find Out If Your Job Will Be Automated.” Bloomberg, 7 July 2017.

³Whitehouse and Rojanasakul (see 2).

⁴VICE Special Report: The Future of Work. HBO, 2019.

⁵“US Department of Homeland Security Cybersecurity Strategy.” US Department of Homeland Security, 15 May 2018.

⁶“SAP Leonardo Machine Learning Foundation.” SAP, 2019.

⁷Bostrom, Nick, and Eliezer Yudkowsky. “The Ethics of Artificial Intelligence.” Machine Intelligence Research Institute, 2011.