AI can describe almost anything with remarkable fluency. Yet description is not the same as understanding. Real understanding grows from lived experience, from being in a living relationship with the world, rather than merely naming it.
We often say that the difference between Gen-AI and we humans is that it does not really understand. But what do we mean by that, exactly?
Take something simple, an apple. We understand an apple through lived experience. We have held one, smelled it, tasted it, bitten into it. We know its weight in the hand, its sound when it falls, its crispness on the tongue. Our understanding is sensory, emotional, and contextual.
Gen-AI has none of this. It has never touched, tasted, or smelled an apple, yet it can write about one with remarkable accuracy. Having absorbed countless descriptions, it can produce a vivid account of colour, texture, and flavour. On the surface, it seems to understand. But what it knows is entirely linguistic, detached from the world that gives those words their meaning.
The difference between our descriptions and those produced by AI lies in meaning, experience, and subjective context.
Our descriptions are rooted in a rich web of emotional, intuitive, and cultural understanding.
We give the apple meaning. It might be a healthy snack, a symbol of education, a fruit of the autumn harvest, or an object that helped Newton think about gravity. Our descriptions can carry emotion or memory: “It reminds me of my grandmother’s apple pie,” or “I love the tart snap of a fresh Granny Smith.” We can even describe an apple through an unusual lens, its sound, a feeling, or a moment, drawing on imagination and intuition.
An AI’s description, while often eloquent, is entirely built from data and patterns. It identifies the words most commonly associated with “apple”: round, red, smooth, and sweet. It can generate these with precision and consistency, but the description has no inner life. It cannot feel delight or nostalgia; it cannot care. Its attention is oriented toward the visual, grounded in vast collections of images and text. Its strength is precision, not perspective.
While our accounts are holistic and meaningful, an AI’s account is technical and objective. The first arises from lived experience, the second from statistical correlation. Both may produce beautiful sentences, but only one carries the weight of being in the world.
Of course, not all of our understanding is grounded in direct experience. Much of what we know about galaxies, microbes, or the past comes through the words of others. In that sense, the line between human and machine is not as absolute as it seems. We, too, rely on patterns of description, sometimes mistaking familiarity of language for true understanding.
Yet we have the capacity to reconnect meaning to experience. We can taste the apple again, recall its scent, feel its crispness, and remember what it is to know something through contact.
Perhaps that is what distinguishes understanding from description: not what can be said, but what has been lived, and what we can feel to matter.
To understand is to enter into a relationship with people, with things, with ideas. AI can imitate the language of relationships, but it stands outside of them. Our task is to keep knowledge alive by remaining in conversation with the world that gives it meaning.
In-person, 7–11 September 2026
Warbrook House, Hampshire, UK
We are living and working in conditions of uncertainty, complexity, and rapid change. This week-long workshop offers a space to practise Conversational Leadership as a shared, lived experience.