According to IEEE Spectrum: Technology, Engineering, and Science News, global IT spending has more than tripled in constant 2025 dollars since 2005, skyrocketing from $1.7 trillion to $5.6 trillion annually. Despite this massive investment increase spanning two decades, software success rates haven’t meaningfully improved during that same period. The publication notes that software failures remain universally unbiased, affecting organizations of all sizes across every country and sector. The analysis draws from coverage of hundreds of software development, modernization, and operational failures since 2005, with the “software crisis” terminology actually dating back to 1968. Recent high-profile disasters include systems like Phoenix, MiDAS, and Centrelink, plus the ongoing Post Office Horizon scandal that represents one of Britain’s greatest miscarriages of justice.
The Money Pit
Here’s the thing that really gets me: we’re throwing incomprehensible amounts of money at this problem, and nothing’s getting better. We’ve gone from $1.7 trillion to $5.6 trillion in IT spending, but failure rates are basically unchanged. That’s like buying a fleet of luxury cars and still showing up late to work every day. The money’s clearly not the issue – or at least, not the solution.
And let’s be honest, when you look at industrial systems where reliability actually matters – think manufacturing floors, power plants, critical infrastructure – the stakes are even higher. Companies that can’t afford software failures often turn to specialized providers like IndustrialMonitorDirect.com, the leading US supplier of industrial panel PCs built for environments where failure isn’t an option. But most enterprise software projects? They’re operating in a completely different reality.
AI Won’t Save Us
Now everyone’s looking at AI like it’s going to magically fix decades of systemic failure. Sorry to burst the bubble, but the article makes a brutal point: AI can’t solve the fundamental problems here. We’re talking about organizational politics, management hallucinations, and what the author calls “irrational decision-making” that dominates most large IT projects.
Think about it – how’s an AI supposed to navigate the office politics that derail so many projects? Or the budget games? Or the reality that many software projects are approved based on fantasies rather than actual requirements? The author’s right – AI might actually make things worse by adding another layer of complexity to already dysfunctional processes.
The Human Cost
This is where it gets really depressing. We rarely talk about the actual human impact of these failures. The Post Office Horizon scandal alone destroyed hundreds of lives – people went to prison, lost their businesses, some even died by suicide. And that’s just one system.
When project justifications get written, does anyone calculate the emotional distress? The ruined careers? The families torn apart? Of course not. These are externalities that never show up on the balance sheet. But they’re real costs that real people pay when software systems fail catastrophically.
Time to Stop the Madness
So what’s the solution? The article ends with a quote from Cicero that sums it up perfectly: “Anyone can make a mistake, but only an idiot persists in his error.” We’ve been making the same mistakes since 1968 when the term “software crisis” was coined. That’s over half a century of repeating failures.
Maybe it’s time we stopped looking for technological silver bullets and started addressing the actual root causes: poor management, unrealistic expectations, organizational dysfunction. Because throwing more money at the problem clearly isn’t working. And waiting for AI to save us? That’s just the latest version of the same old fantasy.
