What Catches Your Mind?

We frequently make quick judgments about what we like—from faces to furniture, colors to cars. But what shapes these preferences?

This digital exhibit reveals the surprising patterns behind our individual choices. Discover what preferences you share with others and which ones are uniquely you.

Start with a simple example of how you make quick judgments in everyday life.

Then, choose a category of things to rate—dogs, cats, faces, or landscapes—and a characteristic to explore. Will you rate serene landscapes, trustworthy faces, friendly cats, or something else?

You’ll rate several images, then a cutting-edge algorithm will generate personalized images based on your preferences. See your results alongside others’ to reveal the ways your preferences are shared between people or uniquely your own.

Images from the Exhibit

Woman in the photobooth with a man standing behind her

Two people pointing at an exhibit together

Man in the photobooth with a man standing behind him

The Science behind the Exhibit

We commonly overestimate how much others will share our opinions - a common bias called the “false consensus effect.” This bias occurs because it's easier to think about the ways people agree with us more than the ways they don't. But the false consensus effect and the assumptions it produces can make it harder to understand how others see the world.

How This Technology Works

Every image in this exhibit is created by artificial intelligence (AI) – none of the dogs, cats, landscapes, or faces are real. This exhibit uses a special type of AI called a Generative Adversarial Network (GAN) to create these images.

How the AI Learns Your Preferences

GANs are good at extracting key features that represent a collection of images without relying on memorizing every single small detail. You can think of it like describing a face with just a few important features, such as the hair color, what the nose looks like, the distance between the eyes, and the shape of the jawline.

As you rate different images, our GAN-AI learns and keeps track of the common features that make up what you think of as “trustworthy faces,” “adoptable dogs,” or “serene landscapes.” This process allows it to create a "blueprint" of your unique preferences.

How the AI Creates New Images Based on Your Preferences

Once the AI understands your preferences, it can create entirely new pictures based on the patterns you like. As you rate more photos, the successive images change in real time to better match your feedback.

Real-World Applications

This technology has many potential applications, from exploring individual-level preferences of consumer products to helping scientists identify the structure and nature of biases.