Old, unreliable software is the $1.5 trillion mystery.
In their haste to create cutting-edge software, companies are piling up an often-overlooked form of debt on their tech balance sheets.
“Technical debt” is the name for it. Hidden behind the veneer of newness and creativity in IT systems are a plethora of antiquated systems that were never meant for their current purpose, as well as hasty modifications. These systems are in dire need of an update.
A plethora of issues, including security breaches, delayed innovation, and system breakdowns, are all symptoms of technical debt. During the peak of the holiday travel season in the United States in late December 2022, it caused more than 13,000 flights operated by Southwest Airlines to be cancelled, leaving passengers and their belongings stranded. Experts agree that it’s a major cause of the numerous software vulnerabilities that have resulted in dozens of cyberattacks over the last year, including attacks on Google, Apple, and Microsoft’s crucial systems.
According to a report from 2022 by a charity backed by the software industry, fixing this technological debt would cost $1.52 trillion. Additionally, the United States loses $2.41 trillion year due to cybersecurity and operational problems, unsuccessful development initiatives, and the upkeep of old systems. The yearly interest paid by the United States government on its national debt is more than 2.5 times that amount. Herb Krasner, a retired software engineering professor from the University of Texas in Austin, claims to have calculated that the national debt has increased to almost $2 trillion.
Like our national debt, technical debt has grown larger over time, skyrocketed during the pandemic, and now threatens to overwhelm us all. Because technology is already causing dangerous situations in our daily lives, it is also similar to our deteriorating physical infrastructure.
A lot of people who work in software development are concerned that the increasing use of AI will exacerbate the situation. The rationale behind this is that technical debt is likely to increase if mechanisms are put in place to facilitate software development and delivery, particularly for less experienced programmers. Instead of spending time optimizing code for maximum efficiency, developers generally prioritize getting things out the door.
The same problem with our shared priorities is the root cause of our infrastructure’s neglect.
One of those invisible concerns that people either realize they have a problem with or don’t know about, and that’s worse, is technical debt,” says Roger Williams, a vice president of research at Gartner. Just as with anything we have at home, it’s easier and cheaper to put things off till tomorrow.
The only way for management to address technical debt is for IT departments to stop taking requests for new, exciting things and instead concentrate on improving the strength, security, and future-proofing of what the company already has.
Some companies may be reluctant to pay for necessary maintenance since they are eager to invest in developing new technical capabilities but aren’t sure it will provide any tangible benefits.
Cast Software’s chief scientist Bill Curtis stated, “The number one thing keeping us from retiring technical debt is the fact that the business wants new capabilities.” The company develops tools to help enterprises comprehend their own complex software systems. “They need to make it a regulation that we allocate a certain amount of time to paying off technical debt, and then they need to make sure we stick to it.”
It is far more difficult to construct new items and switch to new firms without such policies.
Picture software as a massive mechanical device. Adding parts to a machine can make it perform more, but it won’t change its fundamental operation. Also, those internals might be working overtime to power all these additional features. Getting a software machine to do new things is difficult as its complexity increases. As a result, technical debt becomes a growing obstacle to efficiency.
For Krasner, the subject of much research, “People are not changing their ways, in terms of technology and processes” is the key. Getting things out the door fast vs building quality in is truly at the heart of the matter.
Worryingly, we frequently depend on the world’s oldest and most august organizations the most, and their technical debt is the worst. Technical debt, for instance, falls disproportionately on the United States’ banking and financial institutions. Cobol, a computer language developed and distributed by the United States Department of Defense in the early 1960s, is currently in use by many organizations’ computer systems.
“Cobol is heavily used in banking systems, but it has been around for a long time, nobody has bothered to document it, and the original developer is likely no longer alive,” Curtis remarks.
International assistance may be necessary to resolve the technological debt of the United States and other countries. Coding rewrites and, occasionally, system re-architecting, play a significant role in lowering technical debt.
Legally, software, AI, and even erroneous data (such in a navigation system) are now considered faulty products according to new EU regulations. Companies will now be able to be sued for faults in any of those areas, rather than only hardware, according to this amendment.
New software tools could also help the country pay down its technological debt. According to Krasner, AI-powered code editors not only facilitate the generation of code that can add to technical debt, but they can also assist in analyzing current codebases to ascertain which parts require updating and in what ways.
Human resource management software provider ADP is one of several businesses utilizing AI to port legacy code (such as Cobol) to more modern languages where skilled programmers are more readily available.
In the past, when businesses owned their own servers, inefficient systems could be easily remedied by purchasing additional racks of computers. According to Ben DeBow, CEO of Fortified, a consulting firm that assists companies with database management and updates, because of inefficiencies, companies are now paying for computing on an as-needed basis in the cloud. As a result, bills are twice as high as they could be if older systems were optimized.
Aside from legislation and AI, a typical method of eliminating technological debt is through a phenomenon that could be described as technical bankruptcy. This occurs when established organizations are overtaken by upstarts who do not have the same reliance on outdated technologies. Startups in the last 20 years that build software for businesses on the cloud, like Salesforce, have been able to avoid the pitfalls of technical debt, which has been a major factor in their success.
But there are other organizations, like the US government, that can’t be easily replaced with new, improved models. In a recent statement, the chief information officer of the United States Department of Defense hinted at the possibility of sending orders to military groups to upgrade and discard antiquated software and hardware systems.
Saving money and making innovation easier are important, but technological debt is more important from a national security standpoint in a world where advanced adversaries like Russia and China can hack into U.S. networks.
