Face it, There is Nothing That Can Replace COBOL
In the world of programming and information technology, there exists a language that, despite being over six decades old, remains a pivotal foundation of the IT ecosystem and by extension, the global economy. This language is COBOL (Common Business Oriented Language), a name that resonates with an unmatched legacy in the corridors of technology.
COBOL’s significance cannot be overstated – it underpins nearly 90 percent of Fortune 500 company operations and is responsible for 85 percent of all business transactions worldwide. An astonishing fact is that COBOL-powered systems facilitate $3 trillion in daily commerce, make 95 percent of all ATM transactions possible, and support 80 percent of in-person credit card transactions. Philip Teplitzky, a seasoned COBOL developer for US banks, puts it strikingly: “The second most valuable asset in the United States – after oil – is the 240 billion lines of COBOL.”
However, the persistence of COBOL is met with considerable challenges. The shrinking pool of developers familiar with the language poses a threat to its continuity, a situation starkly highlighted during the COVID-19 pandemic as state governments scrambled to find COBOL-skilled technicians for overwhelmed unemployment systems. Moreover, the infrastructure COBOL runs on – mainly mainframes – are notoriously difficult to modernize, presenting a bottleneck for adopting contemporary business practices like mobile integration.
Maryland’s Information Technology Secretary Katie Savage has voiced concerns, labeling COBOL “a significant operating risk” that necessitates an upgrade from both security and workforce development angles. Yet, despite these compelling arguments for moving away from COBOL, transitioning to a modern platform is easier said than done.
Mainframes are still esteemed for their robust security and resiliency, and the sheer volume of operations they handle cannot be easily replicated on newer systems without immense cost. Businesses and governments find themselves in a difficult predicament, weighing the sunk costs of their current systems against the hefty price of modernization. Furthermore, the interconnectivity between COBOL and modern software tools provides a lifeline that keeps its use viable, including possibilities of cloud migration.
It appears that COBOL has no clear successor. While there are more current programming languages like Java or C#, neither governments nor corporations are taking definitive steps to adopt them wholesale. A glimmer of hope appeared with IBM’s announcement of a generative AI tool designed to facilitate COBOL to Java translation, yet this too is not without its caveats. Skyla Loomis of IBM cautions that the approach, requiring developer intervention and being in its infancy, is not a magic bullet.
Gartner’s Distinguished Vice President and Analyst Arun Chandrasekara further tempers enthusiasm by emphasizing the nascent stage of AI in such applications and the lack of validation for IBM’s ambitious tool. The message is clear: while experimentation with AI in bridging the gap left by diminishing COBOL expertise is promising, it is premature to declare it a definitive solution.
The challenges in replacing COBOL, along with its ingrained presence in the global financial infrastructure, highlight a peculiar conundrum. On one hand, the necessity for modern, agile, and secure systems in the digital age is undeniable. On the other, the enduring legacy and operational muscle of COBOL render it a formidable foundation that can’t be easily forsaken. For now, it seems, we must face the reality: there is nothing that can replace COBOL – at least not yet.
