Main image of article How Tech Pro Retirements Will Leave a Knowledge Gap in Tech

Tech is a dynamic industry, constantly evolving through new platforms, technologies, and programming languages. While many tend to focus on programming language changes and which IDE is the new darling of the tech industry, there are a few larger issues that need to be addressed. An aging workforce means there will be retirements en masse, leaving a dearth of tacit knowledge about legacy platforms, code, and more.

Moreover, artificial intelligence (AI) might be the impetus for retirements. We already know AI is responsible for layoffs across the tech industry. As AI matures, it’s likely to continue to squeeze out technologists, and many of them may choose to retire early rather than wait for AI to take their jobs.

As companies remain mindful of the bottom line, we’re left to ask what impact AI will have on older technologists, and how it might reframe the career paths for younger tech pros. To find out, we spoke with several experts to discover what might be next for us in this AI-first world of tech.

Summary

“Older tech professionals often have experience with ‘lower-level’ programming languages that younger generations are less familiar with,” Sylvain Kalache, Head of AI Labs at Rootly, tells Dice. “This reflects the evolution of software development, and I believe won't be any different for what's to come. As new layers of abstraction have been introduced, the need to interact directly with the hardware has decreased for most developers.

“Early programmers worked directly with punch cards and machine code. That transitioned into assembly language, and eventually into early high-level languages, such as FORTRAN and COBOL. Later came C and C++, which are still in use today but are less commonly taught as first languages. In recent decades, higher-level programming languages like Java, Python, JavaScript, and Ruby have gained dominance, particularly with the rise of web and mobile development.

“Now, with the rise of AI tools and large language models, we're entering a phase where natural language prompts may increasingly supplement or even replace manual coding for some tasks. I would say that older talent will likely know more about any programming languages than the younger generation. The younger generation will be more skilled in prompting machines to do the programming.”

There is a rising knowledge gap as experienced IT workers retire,” says Sergio Oliveira, Chief Technology Officer at DesgnRush. “This is especially true for fields that still use archaic programming languages like COBOL, Fortran, and other older versions of Java. Unfortunately, these systems remain vital for the government, healthcare, and finance; however, few people know how to keep them running. AI and automation can help with some things (they prove to us every day), but they can't ever entirely replace the deep, nuanced expertise that comes from years of working with these systems.  People are mistaken in supposing that AI would make older people retire early. It's not AI taking over; instead, the rapid advancement of new technologies and the burnout of people are driving experienced workers to leave.

“To remedy this, businesses need to put mentoring and passing on information at the top of their lists. Engineers who are leaving the field have a unique form of tacit, experiential knowledge. AI technologies can aid with certain activities, but they can't make decisions based on the situation like experienced professionals can. Companies need to do succession planning and strategic training long before workers leave to make sure that the intellectual capital is passed on. Hybrid systems will be highly vital to keep this balance between current efficiency and old-fashioned knowledge. Younger computer pros work with senior experts in these systems.”

One key issue for many applications is the reliance on one or two senior developers who built systems that only they understand,” adds Lee Brewington, Senior Software Developer at Atiba. “This is especially common in small to medium-sized businesses with constrained budgets or fast-moving needs. Often, the applications are poorly documented, and the lead developer is the only person with a comprehensive understanding of the system.

“This creates strain on both the company and the individual looking to retire, as long-tenured professionals may hesitate to leave the business in a vulnerable position. As a result, they may stay on longer than planned or enter into partial retirement. These applications are often custom built or hosted on legacy platforms which younger developers have no experience and little desire to learn.”

“LLMs are impressive in their breadth,” adds Kalache. “They have been trained on decades of code, including many legacy and niche languages. They can be great learning tools, helping younger engineers get started with systems they have never touched. However, they are not yet sharp enough to fully replace the deep, contextual knowledge held by experienced engineers. They can generate code and explain syntax, but they often miss edge cases, struggle with non-standard implementations, and lack awareness of the specific quirks of old production environments.

“So today, the answer is ‘no.’ AI can assist, but not fully bridge the gap. That said, if the pace of improvement continues, it is likely that in a few years, the answer will shift to yes, at least for a meaningful portion of legacy maintenance and modernization work.”

“I'd say ‘no,’ AI isn't forcing senior talent into retirement sooner than expected, I'd say it's the opposite – they will thrive with this tech,” Kalache tells Dice. “There are a lot of layoffs lately, and I think it's multiple trends converging.

“Post-COVID over-hiring adjustments are being resolved. Companies hired too many people at very high salaries during the COVID pandemic. We are seeing a correction. For the last 2-3 years, the tech industry has faced a recession, with VC capital drying out and no IPOs happening, leading to a largely cooled-down market. Startups often shut down, and larger companies tend to be more conservative with their engineering budgets.

“Most companies are asking their engineers to do more with less by using AI. This means that talent that is not able – or willing – to adopt AI to speed up their work won't last. But, quite frankly, I don't believe this is happening for experienced talent. On the flip side, I believe AI is VERY beneficial to experienced talent who can make great use of these tools. Because they know what "good code is, they can ensure that these AI tools are doing it right, versus junior talent who don't know any better.”