Imagine this: every day, $3 trillion worth of transactions are handled by a 64-year-old programming language. Yes, you read that right—$3 trillion. We’re talking about banks, insurance companies, government agencies, healthcare systems, you name it. And that old code? It’s called COBOL (Common Business-Oriented Language), a programming language that your parents’ parents might have worked with. COBOL is still kicking, despite the fact that most schools and universities stopped teaching it years ago. It’s kind of like relying on a flip phone in 2024—unthinkable, but it’s happening.
Here’s where it gets even crazier. COBOL isn’t just lingering in obscure places; it’s actually a major player in industries like banking and finance. According to reports, 43% of all banking systems are still running on COBOL, handling nearly 95% of ATM transactions and 80% of all in-person credit card transactions in the U.S. COBOL is quite literally the engine behind the money that moves around the world every day.
So, what’s the problem? Simple: no one knows how to use it anymore. It’s old, clunky, and coding in COBOL feels like writing a high school essay—it takes forever and is way too detailed. And since universities stopped teaching it long ago, the number of people who can maintain and fix COBOL systems is shrinking by the day. The folks who once ruled the COBOL world are now retiring, and we’re running out of these “COBOL cowboys.” That leaves us in a tricky spot: tons of systems still depend on this ancient language, but there aren’t enough developers left who know how to work with it.
IBM’s Big Plan: Can AI Save the Day?
Enter IBM, the tech giant with a plan to modernize this old code. Their idea? Instead of relying on a shrinking pool of human programmers, why not use artificial intelligence to give COBOL a facelift? They’ve built a fancy AI tool called watsonx, which is like a super-powered code assistant. Think of it as ChatGPT for old-school programming. You feed it chunks of COBOL code, and it can help translate that into a more modern language like Java. Of course, it’s not as easy as pushing a button and watching it work its magic, but it sure does cut down the time and effort needed to reprogram.
IBM’s watsonx isn’t about replacing human developers, though. According to Skyla Loomis, IBM’s Vice President of Z Software, this AI tool works with developers to streamline the process. Sure, it can handle 80-90% of the conversion work, but human developers still need to step in and make a few tweaks to ensure everything runs smoothly. It’s more of a productivity boost than a full-on replacement for developers.
Keri Olson, another IBM exec, explains that the process is thorough and complex. Before any code gets translated, IBM and the client need to fully understand how the existing system works—its data flow, dependencies, and all the little details. Once that’s done, they start breaking the code down into smaller, manageable pieces. This is where the magic happens: COBOL gets turned into Java, one bit at a time, with watsonx assisting along the way.
The Skepticism: Is AI the Real Deal?
As exciting as it sounds, not everyone’s on board just yet. If you’ve been around the tech block a few times, you might remember Watson Health, IBM’s previous foray into AI for healthcare that didn’t quite live up to its grand promises. So, some experts are understandably cautious about watsonx. Arun Chandrasekara, a distinguished analyst at Gartner, isn’t ready to pop the champagne just yet. He points out that IBM hasn’t published any case studies to prove this AI-powered COBOL conversion tool works as well as they claim.
Even IBM admits they’re still in the early days with watsonx, and it’s going to take time to show solid results. But they remain optimistic. Olson highlights that IBM’s experience with Z computing and AI is setting the stage for something big. If watsonx lives up to the hype, it could change the way companies modernize not just COBOL, but all kinds of outdated code.
The Future of AI and Code
If IBM’s AI-assisted code conversion works as they hope, the implications could be massive. In fact, it’s not just about COBOL. Gartner predicts that by 2028, the partnership between AI assistants and human developers could reduce the time needed to complete coding tasks by 30%. That’s huge. Imagine what it would mean for the tech world if AI could help programmers churn out code faster, more efficiently, and with fewer headaches. We might be closer than we think to a future where AI is just another tool developers use to get the job done.
But of course, we’ve got to wait and see if this all pans out. Will IBM’s watsonx revolutionize the way we handle ancient, crucial programming languages like COBOL? Or will it fizzle out like some other AI projects in the past? Only time will tell. But one thing’s for sure: we’ll be watching closely as this new chapter in tech unfolds.
