I was preparing a talk the other day — about AI, about healthcare, about the usual things a CTO is asked to talk about — when I put two images next to each other on a slide. The first was a Sumerian clay tablet, roughly five thousand years old. Cuneiform pressed into wet earth by a reed stylus, then baked hard by the sun. Each tablet was a unique object, handmade. Only a small elite of trained scribes could read it — perhaps two percent of the population. Knowledge was scarce. It was locked up. And it was power.
The second image was a map of all human scientific knowledge — a network of coloured dots representing every discipline from archaeology to biochemistry, connected by lines showing how researchers move between them. It was built from a billion interactions with scholarly journals. It was published in 2009. Already seventeen years old.
The contrast is obvious, and I almost moved on. From a world where few could read anything, to a world where everything is readable. From clay to cloud. From scarcity to abundance. The story writes itself.
But then I paused. Because something didn’t fit.
We tell ourselves a clean narrative about knowledge. It goes like this: once, knowledge was scarce and held by the few. Then came the printing press, then literacy, then the internet. Now, knowledge is everywhere and belongs to everyone. Problem solved.
Except it isn’t. Not really.
Because alongside the explosion of knowledge came specialisation. And specialisation, quietly and without anyone quite noticing, recreated the original asymmetry — just in a different shape.
Consider: a cardiac electrophysiologist understands the electrical mapping of the human heart in a way that perhaps 0.001 percent of the world’s population can follow. A maritime lawyer knows the jurisdictional nuances of international shipping law. A quantum physicist works in a language that even most other physicists find opaque. These are not rare examples. This is the normal condition of modern knowledge.
We went from a world where few could read anything to a world where everyone can read, but few understand any one thing.
The proportion of people who grasp a given piece of specific knowledge may not have increased since the Sumerians. It may even have decreased. The tablet has been replaced by the paper, the paper by the database, the database by the model — but the fundamental asymmetry persists. The person who holds the knowledge still holds the power. The farmer who needed the scribe to read the temple records is not so different from the patient who needs the specialist to interpret the scan.
The shape of the asymmetry changed. Its effect did not.
I’ve spent thirty years in technology, and for most of that time I assumed that the problem was access. Give people access to information, and the asymmetry dissolves. That was the promise of the internet, of open data, of Wikipedia. And it was partly true — but only for knowledge that is already accessible to the general reader. For specialised knowledge, access alone is not enough. You can read a research paper on arrhythmia and still not know what it means for your heart.
What AI changes is not access. We already had that. What AI changes is comprehension on demand.
For the first time, you can encounter a piece of specialised knowledge and ask for it to be explained in terms you understand. Not simplified to the point of uselessness, but translated — contextualised, personalised, made conversational. You don’t need five years of medical school to understand your diagnosis. You don’t need a law degree to read your contract. You don’t need a background in finance to understand your pension statement.
But I have learned to be cautious with “for the first time.”
The printing press looked like it would dissolve the asymmetry. It didn’t — it moved it. The internet looked like it would dissolve it. It didn’t — it moved it again. Each time, the distance between the person who holds the knowledge and the person who needs it reappeared in a new shape, wearing new clothes, speaking a new language.
Maybe comprehension on demand breaks the pattern. Maybe it just shifts it — from who holds the knowledge to who knows how to interrogate the explanation. From the scribe to the specialist to whatever comes next.
What does seem to persist, across every version of the asymmetry, is that the distance is always shortest for the curious — the ones who observe, consider, and ask whether what they’ve been told matches what they see. Every question comes from a place of less understanding. That’s not a weakness. It’s the only posture that has ever closed the gap.
I keep coming back to those two images. The clay tablet and the knowledge map. Five thousand years apart. The asymmetry present in both.
Whether what is happening now actually dissolves it, or merely moves it again — I don’t know. But I think it is the right question. And I think the only honest way to answer it is to watch, carefully, and see what the asymmetry is wearing next time it appears.