The shortsightedness of replacing the human workforce with AI

It doesn’t seem that long ago that the IT industry was sounding the alarm over a chronic shortage of software engineers, database administrators (DBAs), and cyber-security professionals. Today, the story has changed tack as people start to question the extent to which those same jobs could be replaced by AI. 

 

While it’s a remarkable turnaround in terms of narrative, for many, it’s the speed at which things have moved that is the real eye-opener. In just a few short years, the focus has shifted from generative AI (GenAI), with its ability to create text, summarise information and even write code, to agentic AI and how it makes decisions and carries out tasks with ever-increasing levels of autonomy.

 

Our society has pivoted as dramatically in the past, almost always due to a dramatic and powerful new technology. From the steam-powered textile mills and the Industrial Revolution, from train transport and steamships and internal combustion engines, from electronic communications like the telegraph to electrical infrastructure, each new technology caused significant instability and disruption to the people of that era. In some cases, these shifts destabilised entire sectors of the economy, supplanting the incumbent with something both better and worse than what existed before. GenAI is no different.

 

 

The benefits of AI to IT teams 

On the face of it, the potential benefits for IT teams are obvious. But disastrously destabilising technology often arrives dressed as good sense. An educated and appropriate call, if you will. After all, agentic AI promises to automate routine administration, commit code faster than ever, speed up troubleshooting, improve collaboration across teams, and reduce the burden of repetitive work.

 

In theory, after setting a careful and deliberate strategy, GenAI allows skilled professionals to spend more time solving complex problems while helping to drive innovation and adding value across the wider business. But as I alluded to earlier, my concern is that many do not agree with this approach. In fact, many business leaders see it as a cost-cutting measure that should be rushed into immediate implementation, a way for AI to replace people rather than make them more productive.

 

In my view, this approach is a mistake. In some cases, it may end up being a grave and costly mistake. For if firms do start to let staff go, these organisations risk losing the expertise, judgement and experience that only humans can provide. For me, it boils down to three core reasons.

 

 

The loss of institutional knowledge

First, one of the greatest strengths of AI is its ability to analyse vast amounts of information. But organisations don’t run on data alone. They also rely on years of accumulated nous and know-how that rarely appears in a database or technical manual. Ask any experienced IT professional why a particular system is configured in an uncommon way, why a workaround exists, or why one approach was chosen over another, and the answer is often rooted in years of ‘first-hand experience’ rooted deeply within their organisational context.

 

It’s the kind of institutional knowledge that’s built over time through solving problems, learning from mistakes, and understanding how the organisation really works. AI can certainly help teams access and interpret information more quickly. But it can’t replace the insight that comes from having lived through major incidents, technology migrations, and changing business priorities. If organisations lose that experience in the pursuit of short-term efficiency, they risk discarding one of their most valuable competitive assets.

 

 

AI that is rendered ineffective

Here’s a historical antecedent: manufacturers tried for twenty years to replace steam after the discovery of electricity. But it never outperformed steam until the engineers learned running their textile looms from a central drive shaft, like with steam, was the problem. Electric turbines, when they were seated beside each textile loom, finally and significantly outperformed steam – because they reengineered the entire process!

 

My second concern is similar and centres on the assumption that AI can operate just like we did in the past, but independently of people. Truth is, despite all the progress to date, it is only as effective as the processes used and the data it receives.

 

Even when good processes are implemented and the data is excellent, AI is still far from infallible. It can misunderstand context, produce inaccurate responses and, as we’ve all come to recognise, hallucinate. It will simply lie. That’s why human oversight remains essential. Yes, AI can accelerate decision-making. But it shouldn’t replace the experience and judgement needed to validate its output.

 

 

The ill-advised dismissal of junior talent

And finally, if organisations do begin to replace people with AI, my fear is that junior roles are likely to bear the brunt of those losses. Worse, young people simply won’t be given the opportunity to get their foot on the career ladder. While that may deliver short-term efficiencies and reduce personnel costs, it risks creating a much bigger problem further down the line. After all, today’s junior engineers are tomorrow’s senior architects and IT leaders. 

 

If we remove those entry-level opportunities, then organisations risk disrupting the very talent pipeline they depend on. With the average DBA already in their mid-50s, investing in the next generation has never been more important. The workforce in many IT disciplines is facing a steep increase in retirements over the next 10 years. Who will be available to take their place?

 

 

Ensuring AI is a partner, not a replacement

To me, these three arguments are a wake-up call for our profession. To use AI simply to reduce headcount in this way is simply shortsighted.

 

A far better approach is one that finds the right balance between human expertise and intelligent automation. Yes, AI should be used. But to alleviate toil, such as removing repetitive tasks, accelerating decision-making, and supporting employees. It should not be used to replace the experience, judgement and critical thinking that career professionals bring to the table.

 

Rather than focusing solely on short-term productivity gains, business leaders should be asking whether they have the right foundations in place to support AI over the long term. Is their data fit for purpose? Are the right governance and safeguards in place? And, perhaps most importantly, are they investing in the people who will be expected to work alongside these technologies?

 

By the way, it’s also up to IT professionals to move with the times as well. If they want to be freed from repetitive tasks and take on more challenging problem-solving, they need to develop new skills to use AI effectively.

 

If there’s one thing we’ve learnt from past technological advances, it’s this: technology alone has never been the foundation of a successful organisation. People have. We would be wise not to forget that.

 

---

 

Kevin Kline is a Database Technology Evangelist at SolarWinds

 

Main image courtesy of iStockPhoto.com and Amorn Suriyan