There’s a quiet assumption we make about knowledge that once it’s created, it can simply be moved. Packaged. Shared. Applied. But you’d find that it doesn’t work that way.

Professor Nevine Labib has spent years watching knowledge travel across countries, institutions, and disciplines. From research labs into policy rooms. From policy into real-world systems. From Cairo to Geneva and back again. And somewhere along the way, she found that something shifts.

“When AI knowledge moves,” she said, “from research to policy to implementation, the breakdown tends to happen in moving data to policy.” That line stays with you. Why? Because it challenges what most people expect. You’d think the problem would be technical. Or theoretical. But it isn’t. It’s translation.

There’s something that happens when an idea leaves the environment in which it was born. In controlled settings, everything behaves. Variables are known. Assumptions are shared. But once that same idea is introduced into a more complex, unpredictable space, it starts to bend. And this is not because it’s wrong. But because it was never built to travel.

AI systems carry this tension in a very visible way. A model trained on clean, structured data begins to fail in bits when exposed to the real world, where data is incomplete, inconsistent, and shaped by human behavior. And then decisions are built on top of that.

“The weakest link we see right now in AI knowledge,” Nevine explained, “is the data quality.”

But you see, even that isn’t the full story. Why? Data isn’t just data. We must remember this. It carries the context of what was measured, what was ignored, and what was assumed to be normal. When that context doesn’t move with it, the data doesn’t just degrade. It misleads.

Still, something holds.

“The core concepts… remain the same,” she said. “It is the focus that changes completely.”

And maybe that’s where the real work is. Not in creating more knowledge. But in understanding what happens to it when it moves.

If you’re building, researching, or applying AI in any form, take a moment this week to ask yourself: where could this break when it leaves my environment?

And if you want to go deeper into how these transitions actually play out across sectors and countries, listen to the full conversation with Prof. Nevine. It will change how you think about “applying” knowledge.