Dimensions Matter
Dimensions Matter
Imagine you're looking for a house. The real estate agent asks, "What kind of house are you looking for? " You reply, "A good one. " That's not enough. So you begin adding details. Three bedrooms. Two bathrooms. 1,800 square feet. Near a metro station. Within your budget. Garden. Parking. Suddenly, the house isn't described by one number. It's described by many. Each new characteristic tells us something new about the same house. Now imagine removing one of those characteristics. Suppose we ignore location. Would two houses still appear similar? Maybe. Now ignore price. Now ignore size. Eventually, very different houses begin to look identical. Why? Because we've lost information. Mathematics calls each independent characteristic a dimension. A house with Price Area Bedrooms exists in a three-dimensional feature space. Add Age Parking Distance from school and it now exists in a six-dimensional space. The object hasn't changed. Only our description of it has become richer. The same idea applies everywhere. A student may be described by Marks. Attendance. Projects. Communication skills. Programming ability. A patient may be described by Blood pressure. Heart rate. Temperature. Blood sugar. Oxygen level. A customer may be described by hundreds of buying habits. Every new feature becomes another dimension. We often imagine dimensions as physical directions. Length. Width. Height. But Machine Learning uses a much broader definition. A dimension is simply an independent piece of information. It doesn't have to be something you can see. A customer's income is a dimension. A movie's genre is a dimension. The sentiment of a review is a dimension. Even the frequency of a word in a document can become a dimension. Now comes the astonishing part. Humans struggle to imagine more than three dimensions. But mathematics has no such limitation. Ten dimensions. One hundred. One thousand. Ten thousand. To mathematics, they are all simply vectors with more coordinates. This is where Artificial Intelligence begins to look very different from human thinking. When ChatGPT represents a sentence... it isn't placing it in three-dimensional space. It represents that sentence as a point in a space with thousands of dimensions. Every coordinate captures a tiny aspect of meaning. No single coordinate means "happiness" or "science. " Together, they form the sentence's mathematical identity. This explains one of the greatest strengths of Machine Learning. The richer the description... the richer the representation. More dimensions allow the model to distinguish between objects that previously looked identical. But they also introduce new challenges. As dimensions increase, the space becomes unimaginably vast. Objects spread apart. Finding neighbours becomes harder. Similarity behaves differently. The mathematics itself begins to change. This phenomenon is known as the Curse of Dimensionality, and it shapes almost every modern AI system. So dimensions are far more than axes on a graph. They are the vocabulary with which mathematics describes reality. Every additional dimension is another question we ask about an object. The better our questions... the richer our representation. And the richer the representation... the more intelligently AI can reason about the world. If AI works in spaces with hundreds or thousands of dimensions, another question becomes inevitable. How can we possibly visualize such spaces? Can mathematics help us understand a world that our brains were never designed to imagine? That takes us to the next chapter. Living in High-Dimensional Space.