This article examines why AI-generated designs tend to be similar, attributing it to the way AI tools learn from existing patterns in their training data. It offers strategies for designers to break away from predictable outputs by using multiple AI tools, injecting specific constraints, and focusing on human creativity in the design process.