Back in the late 90s, while working as a mainframe application programmer at a local council, I witnessed an interesting conversation between the Director of IT and a Systems Architect about the choice of database system provider. A compelling argument was made to move away from the incumbent Oracle platform. The cost-benefit case was strong, and the competitor's future roadmap made the decision seem straightforward. Yet, the decision was made to stay with the legacy system. "I have to look after my development team.
They're invested in the technology stack. " In the years that followed, I saw more and more examples of technology choices being made for reasons that weren't purely business-driven. Senior leaders dragging their feet; and in too many cases-far too many to be exceptions-deliberately derailing and undermining initiatives and projects. As with the decision to continue with Oracle, and the dozens of examples that followed, I could empathise with the actions taken. If I were in the same position, perhaps I might have done the same.
Wait… that's not true. I wouldn't have done the same. For me, technology is a powerful tool-an enabler-something we as IT professionals deliver for our customers to solve real world problems. The moment we let bad design choices, poor processes, or weak execution dictate what and how we deliver IT, we have failed. Over the years I've worked with hundreds of developers, designers, testers, architects, and product managers. Some were brilliant, some poor, most tried their best. But a fair number made decisions-or forced outcomes-that served their own agenda ahead of those of the key stakeholders.
When Forbes recently published their article on Why 95% of AI Projects Fail, cynics were quick to "send all" to reinforce their anti-AI stance. I received five emails and two LinkedIn posts in just three days, as if I were somehow the leader of the Flat Earth Society being presented with proof the world is round. Of the seven senders of these "truth bombs, " five were developers, one was a Creative Designer, and the other a Project Manager. None quoted the article title in full (most likely having copied the same LinkedIn post doing the rounds).
And none appeared to have read the article itself, or they'd have come across the crucial section: "… a n d w h a t businesses can do about it. " The article makes two critical points: 1. Alignment matters more than algorithms in preventing AI pilot failure. Companies already struggle to keep their arrows pointing in the same direction. Strategy lives in PowerPoint, but marketing runs in one lane, sales in another, and operations somewhere else entirely. 2. Technology change is cultural change. The MIT report highlights the rise of "shadow AI.
" Over 90% of surveyed employees already use personal AI tools like ChatGPT at work, yet only around 40% of companies have o'icial licences. This gap shows how disconnected many corporate initiatives are from how people actually work. Cultural friction is often what sinks technology projects. IT departments worry about performance and risk. HR worries about culture but isn't trained in process integration. Line managers are caught in the middle. Without deliberate attention to culture, adoption collapses-no matter how capable the software.
I suspect the majority of root causes of AI project failures mirror those of IT projects in general: poor design, poor execution, poor alignment. This is the same reason almost 90% of IT projects failed to deliver significant business value long before ChatGPT was even a twinkle in Altman's eye. Here's the thing: whether you are an individual, a small business, an enterprise leader, or even a modern-day Luddite, the decision to embrace AI-or any emerging technology-has always been a complex calculation of what you personally stand to gain or lose.
That alone should make you a little more invested in the decisions you make.