Close your eyes for a moment and picture a house. What did you see? Perhaps a square building, a triangular roof, a door in the middle, a few windows, and maybe a small garden outside. 

Now here is the more interesting question. Why did you see that house? I never told you where it was. I never said it had to be made of bricks. I never said it had to stand on land. It could have been underwater, hanging from a cliff, floating above Colombo, built inside a giant tree, or shaped like something no one has ever called a house before. Yet many of us probably imagined something strangely similar.

That is because our minds are built around patterns. From childhood, we see certain images repeated so often that they become our default. Houses look a certain way. Offices look a certain way. Leaders behave a certain way. Success follows a certain path. 

Creativity often begins when we notice the default and ask whether it is the only option.

This is also how large language models work. Artificial Intelligence (AI) systems are trained on enormous amounts of information. They recognise patterns, relationships, styles, and structures. When we ask a question, they predict what is most likely to come next. That is why they are so useful. They can generate a familiar answer quickly, clearly, and often impressively. But the most likely answer is not always the most creative one.

If 10 people ask an AI tool to design a house using the same simple prompt, they may receive 10 different images. But look closely and the ideas may still feel similar. Different colours. Different windows. Different furniture. Yet the thinking underneath may remain the same. The house changes, but the pattern does not.

This is one of the quieter risks of using AI for creative work. It can help us produce more ideas, but if we all ask the same questions in the same way, it may also lead us towards the same answers. It is like giving a hundred chefs the same ingredients and the same recipe book. The meals may look slightly different, but eventually everything begins to taste familiar.

The problem is not AI. The problem is what we bring to it. We often ask the tool to be creative before we have done any imagining ourselves. We type, ‘Give me a unique campaign idea,’ or ‘Design something innovative,’ and expect originality to appear from a vague request. But originality needs material. It needs memory, culture, emotion, contradiction, and lived experience. AI can connect dots incredibly quickly, but humans still need to choose the dots.

A more interesting house might begin with a more interesting question. What if the house was designed around the sound of rain? What if every room represented a different stage of someone’s life? What if the house could move depending on the weather? What if it was built for a family that rarely sat in the same room but still wanted to feel connected? Suddenly, the task changes. You are no longer asking for a house. You are asking for a world.

That is where human imagination becomes powerful. We can bring details that do not exist in general data. The smell of tea in a kitchen at six in the morning. The way monsoon rain sounds on a tin roof. The memory of a grandparent’s home. The small design choices that make a place feel safe rather than simply beautiful. AI has access to patterns. You have access to experience. And experience is where originality often begins.

This matters far beyond art or design. In business, marketing, education, technology, and leadership, the same principle applies. If we all use the same templates, ask the same questions, and follow the same suggestions, we may become faster without becoming better. The skill of the future may not simply be knowing how to use AI. It may be knowing how to challenge the first answer it gives you.

Ask what is missing. Ask what feels too familiar. Ask what someone from a different background might see. Ask what only you can add. Think of AI as a telescope. It can help you see further, but it cannot decide where you point it. If everyone points in the same direction, we all discover the same sky. Creativity begins when someone turns the telescope somewhere unexpected.

So the next time you work with AI, do not just ask it for an idea. Bring one. Bring a strange memory. Bring a local detail. Bring an unusual combination. Bring a question that has not been asked in quite that way before. Because the future will not belong only to those who can generate the most. It will belong to those who can imagine beyond the most probable answer.

Now close your eyes again and picture a house.

This time, try not to see the one you have seen before.