After eight weeks at Kellogg studying AI transformation, the certificate landed today.
I learned a lot. More than I expected. The technology stack, the frameworks, the genuine capability of what generative and agentic AI can already do.
But the insight that landed hardest came from a case study a classmate brought into our cohort discussion a few weeks in.
GE spent $4 billion over six years building Predix. It was supposed to be the world's most powerful industrial AI platform. The technology worked. The platform failed. Not because the AI wasn't good enough. Because the people it was built for never bought into it. Customers didn't understand it. GE's own salesforce couldn't explain it. Nobody had started with actual human needs. The platform was eventually sold off.
GE didn't fail because it underestimated the technology. It failed because it overestimated it.
Across 66 students on the programme, representing multiple sectors and geographies, we each ran a diagnostic on our own organisation. The pattern was similar every time. Hardly anyone found AI capability to be their primary constraint.
It was usually one of three things: a problem that hadn't been defined sharply enough to solve, data that existed but wasn't in a usable state, or people who hadn't been brought along on the journey.
I went into the course half-expecting the technology to be the limiting factor. It almost never was.
Where most AI strategies quietly come apart isn't in a failure to secure the most sophisticated models. It's in a failure to figure out how to bring human judgement, human context, and genuine buy-in into the equation alongside the technology. Solving real problems for real people, not just building impressive platforms.
The technology is rarely the sole constraint to success. The human side is where it gets hard, and where the real opportunity sits.
The question I'll take into every conversation about AI from now on isn't "which model?" It's "how do humans feature in the thinking?"
Thank you Mohanbir Sawhney, Ariana Smetana, David Kolb Tayo Olabumuyi, P.E., Edmund Sefton and the rest of the faculty at Northwestern University Kellogg School of Management for such a thorough and thought-provoking course.