Will AI replace IT first? Numerous IT professionals in my office express skepticism, claiming that AI is not as effective as it is said to be. I have utilized AI to resolve various IT issues/challenges.
8 replies (most recent on top)
@gb and @hz, Will my Cobol coding skills be useful? renders: Yes, your COBOL skills will likely remain valuable for quite some time, for several reasons:
High demand, shrinking supply - There's a critical shortage of COBOL programmers as the existing workforce retires, while billions of lines of COBOL code still run critical systems in banking, insurance, government, and other sectors. This creates strong demand for COBOL expertise.
AI limitations - Current AI tools struggle with the complexity and nuances of legacy COBOL systems. They often can't accurately handle business logic, dependencies, and the institutional knowledge embedded in decades-old code. Human expertise is still essential to validate, guide, and correct AI-generated conversions.
Hybrid approach - The most realistic scenario is that COBOL developers will work alongside AI tools rather than be replaced by them. You'd use AI to accelerate migration projects while applying your expertise to ensure accuracy, handle edge cases, and make critical decisions.
Long transition period - Even with AI assistance, migrating massive COBOL codebases is a multi-year (or multi-decade) undertaking. Many organizations may choose to maintain their COBOL systems rather than risk costly migrations.
Your COBOL skills combined with knowledge of modern languages and AI tools would be an especially powerful combination in this evolving landscape.
Now isn't the mide/mode/micky mouse codebase the same as having a "Cobol codebase"? With all the talent being kicked out the door.... And who cares about that risk anymore!
AI is not effective because of the person supposedly leading it or was leading it T Kerrins. Now it’s been handed to SVB who has little clue as well. Bridgit has no grasp on anything and only like the breeze blown up her backside.
@OP AI doesn’t have to BE effective to cause people to lose jobs. WF needs to cut more staff so the “rationale” is AI replacement is baked into the budget. Whether AI is real or effective doesn’t even matter in the short term
AI is great for reporting, being a search engine for documents spread across multiple systems, spell checker and grammar checker.
That’s about it.
@gb People can be COBOL programmers for an entire career still. That stuff is going nowhere. Like 25 yrs ago or more people were talking about migrating all that COBOL to java or something -- here we are with a ton of systems still running COBOL.
IMO, this AI stuff is just the latest hype, just like Agile and Cloud. We are not likely to see anything that really makes a dent for years, if that. This will be b/c of the technology + how utterly incompetent we are at doing these sorts of things.
It is going to be interesting to see IT bully out AI as they have bullied out team work, CD/CI, agile transformation, task automation, cloud migration etc. Can't wait to hear the excuses management makes. For those quietly quitting this organization consider the cost of not learning and not using this transformational technology. Hey you can only remain a Cobol programmer for so long .
Boring
AI can't replace a human for anything but the most basic requests. It's basically a more powerful VRU that can handle more than DTMF input. It's good at text generation of basic functions, but without significant hand-holding, it gets lost rather quickly.
Try using it for anything more complex than wordsmithing some boilerplate communication or summarizing information you could search up yourself and you'll quickly see it's limitations.
It's effectively a search engine with better pattern recognition. Which makes sense when you realize it's statistical probability applied to text processing, combined with consensus approval.