Industry Insights
2023: The Year AI Changed the Conversation
In November 2022, ChatGPT went mainstream. By January 2023, every client I spoke to wanted an AI strategy. By March, half of them had started one. By June, most had stalled — not because the technology didn’t work, but because they couldn’t articulate what problem it was supposed to solve.
Sound familiar? It should. It’s the same gap I’ve been writing about for five years.
The Hype Was Extraordinary
I don’t say that dismissively. The capabilities were genuinely remarkable. Large language models could draft documents, summarise reports, generate code, and hold conversations that felt substantive. For the first time, technology demos didn’t require an asterisk. The thing on screen actually did what the presenter said it did.
The enterprise response was predictable: excitement, followed by a scramble to find use cases, followed by pilot projects that proved the technology worked, followed by a quiet realisation that nobody had thought about how it would fit into actual operations.
The Confusion Was Also Extraordinary
Here’s what I observed across a dozen conversations in 2023: organisations were starting with the solution and working backwards to find a problem. “We need to use AI” became the brief. Not “we have a customer service bottleneck that AI might help with” or “our analysts spend forty percent of their time on data preparation that could be automated.” Just: “AI. Go.”
This is the delivery anti-pattern I identified in 2019, wearing a new outfit. The technology changes every few years — cloud, low-code, AI — but the mistake is always the same. Jumping to a solution before understanding the problem. Building before analysing. Assuming that better tools will compensate for unclear thinking.
They won’t. They never have.
The Organisations That Got It Right
The minority of organisations that made genuine progress with AI in 2023 did something unfashionable: they started with the problem. They mapped their processes, identified the points of friction, quantified the cost of manual effort, and then asked whether AI was the right intervention.
Sometimes it was. Automated document classification for a financial services firm saved hundreds of analyst hours per month. Natural language search over a knowledge base transformed how a logistics company’s support team operated. These were real, measurable outcomes.
But in every successful case, the AI implementation was preceded by thorough analysis. The teams knew what they were building and why. The technology was the last decision, not the first.
The Mirror
2023 held up a mirror to the enterprise delivery world, and the reflection wasn’t flattering. The AI hype revealed that most organisations still didn’t have a reliable way to move from “we have a problem” to “here’s the right solution.” The frameworks existed for building. They existed for managing projects. But the critical thinking that should connect business need to solution design — the analysis — was still ad hoc, inconsistent, and undervalued.
I’ve been saying this for five years now. The specific technology changes. The fundamental delivery problem doesn’t.
Something Brewing
There’s a reason I mention all of this in the past tense, even though AI is still the dominant conversation in every boardroom. It’s because 2023 clarified something for me personally.
For fifteen years, across financial services, manufacturing, logistics, energy, and government, I’ve been accumulating notes, frameworks, heuristics, and hard-won lessons about what makes enterprise delivery work. Not project management. Not agile ceremonies. The actual analytical thinking that happens — or should happen — between identifying a problem and designing a solution.
In 2019, I noticed the pattern. In 2020, the pandemic proved it. In 2021 and 2022, I saw it compound. In 2023, with AI as the latest catalyst, it became undeniable.
Those fifteen years of notes started taking shape as something more structured in 2023. Something worth writing down properly. Something that might be worth sharing beyond individual client engagements.
More on that when it’s ready.