Taught by Professor Mark Esposito, PhD, I completed “AI in Business: Creating Value with Machine Learning.” Thinking back, it’s fascinating to see what has changed… and what hasn’t.
What’s changed?
1. AI has gone from experimentation to expectation
Before I boarded a plane to Boston, ChatGPT didn’t exist. Generative AI hadn’t entered boardrooms. Microsoft Copilot wasn’t in every software demo. Most organisations were still asking whether AI would matter. Today, the conversation has shifted from “Should we use AI?” to “How do we scale it responsibly?”
The challenge is no longer finding use cases. It’s embedding AI into the way organisations actually operate.
2. ‘Human in the loop’ matters more than ever
As AI becomes more capable, the importance of human judgement hasn’t diminished — it has increased. The organisations seeing the greatest value aren’t removing humans from decisions; they’re redesigning work so people provide the oversight, context and accountability that AI simply can’t.
3. Successful AI programmes are organisational change programmes
One lesson has become impossible to ignore. The companies making AI stick don’t treat AI as a technology rollout. They treat it as an organisational transformation. Recent Harvard Business Review research continues to reinforce that scaling AI depends on leadership alignment, governance, capability building and adoption — not simply deploying another tool.
4. Governance has become a board-level conversation
Four years ago, governance felt like a future consideration. Today it’s one of the first questions executive teams ask. Who owns AI decisions? What risks are acceptable? How do we ensure consistency across the organisation?
What still stands?
1. Value creation still comes first
One of my biggest takeaways from Harvard remains unchanged: the value proposition sits at the heart of every successful AI initiative. Who is the end user? What are they trying to achieve? Why does solving this problem matter? If you can’t answer those questions, no amount of sophisticated technology will compensate.
2. We still use “AI” as a catch-all phrase
AI isn’t a single technology — it’s an umbrella term covering everything from machine learning and deep learning to neural networks and generative AI. Each has different strengths, limitations and commercial applications. The organisations getting the greatest value aren’t necessarily the ones investing the most. They’re the ones asking the better questions.
3. The Turing Test is still remarkably relevant
One concept that has stayed with me since Harvard is Alan Turing’s famous test for machine intelligence. As a Manchester resident, it makes me particularly proud that our city continues to honour his legacy — not least through the recent establishment of No.10 North. More than 70 years after Turing proposed his famous thought experiment, his central question still challenges us: not whether machines can think, but how we define intelligence in the first place.
4. Every AI decision involves trade-offs
Even in 2022, the World Economic Forum was discussing the need to rethink policy to enable more agile governance. That tension hasn’t disappeared. Organisations continue to balance innovation with privacy, speed with control, and experimentation with responsible governance.
My biggest reflection
The technology has evolved at extraordinary speed. The fundamentals of successful transformation haven’t.
The organisations creating the greatest value with AI aren’t necessarily those with the newest tools. They’re the ones with the clearest strategy, the strongest leadership alignment, and the discipline to turn isolated pilots into sustainable organisational capability.
That’s as true today as it was when I walked into a classroom in Boston four years ago.
I’d love to hear your perspective: what’s the biggest lesson AI has taught your organisation over the last four years?