The Science of Beauty: How Your Brain’s Energy Budget Shapes What You Like

Beauty may be in the eye of the beholder, but the brain just wants to take it easy.

What if the reason you find some images pleasing and others uncomfortable has less to do with personal taste and more to do with biology? A study from the University of Toronto reveals that aesthetic pleasure may be hardwired into our visual system as a preference for efficiency. The less metabolic energy the visual system seems to spend on an image, the more people tend to enjoy looking at it.

Using both artificial intelligence models and brain scans of human viewers examining nearly 5,000 photographs, researchers discovered a clear pattern. Images that people rated as more aesthetically pleasing triggered less neural activity in visual processing regions. The relationship held true whether the measurements came from silicon or gray matter.

“Energy efficiency is a major driving force in the evolution of organisms,” the researchers explain in their paper published in PNAS Nexus. The visual system alone consumes about 44% of the brain’s energy, making efficiency a constant pressure on how we perceive the world around us.

Computer Models Reveal the Pattern First

The research team, led by Yikai Tang and colleagues, started by feeding 4,914 images of objects and scenes into an artificial neural network called VGG19. This computer system was designed to recognize visual patterns similarly to human vision. They measured how many artificial neurons lit up as the network processed each image, treating this as a stand-in for metabolic cost.

Images that human observers rated as more enjoyable actually required fewer active neurons in the trained network. When they compared this to 1,000 untrained versions of the same network, only 18 showed a stronger inverse relationship than the trained model. In other words, learning to recognize images efficiently seemed connected to what humans find pleasing.

Human Brains Show the Same Energy-Saving Preference

Moving from artificial systems to biological ones, the researchers examined brain activity data from four people viewing 4,916 images during brain scans. They used functional MRI, a technique that tracks blood flow in the brain. When neurons work harder, they need more oxygen-rich blood. By measuring these blood flow changes, scientists could estimate how much energy different brain regions were spending.

The pattern matched what they’d seen in the computer model. Lower-level visual areas, the parts of the brain that first process what your eyes see, showed modest negative correlations between brain activity and aesthetic ratings. Think of these regions like the first workers on an assembly line, handling basic tasks like detecting edges and colors.

Higher-level regions specialized for recognizing faces and scenes displayed much stronger relationships. The parahippocampal place area, a brain region that helps you understand spatial layouts and environments, showed the most pronounced effect. These are like the quality control inspectors at the end of the assembly line, making sense of the complete picture.

The brain consumes 20% of the body’s energy, with nearly half of that going to vision. This creates evolutionary pressure to process visual information as cheaply as possible without sacrificing accuracy. Think of it like your phone’s battery, if looking at certain images drains the battery faster, your brain might have evolved to find those images less appealing.

The research team proposes that aesthetic pleasure might function as a quick emotional readout that guides behavior without requiring conscious calculation. Just as pain signals tissue damage and hunger signals caloric need, visual pleasure may signal efficient neural processing. Your brain essentially says “I like this” when it can process something easily.

Why Some Images Feel Uncomfortable

This framework helps explain several puzzling phenomena. Studies have shown that faces closer to the average of their category are judged more attractive. Familiar objects and scenes become more pleasing with repeated exposure. Visual patterns with certain spatial frequencies feel comfortable while others create discomfort.

All of these effects could stem from how efficiently the brain can process them. Average faces might require less work than unusual ones because the visual system has encountered similar patterns more frequently. It’s like the difference between reading a word in your native language versus decoding an unfamiliar foreign word. Familiar scenes become easier to encode over time, like a route you drive every day requiring less conscious attention than a new road.

Certain visual patterns might trigger excessive neural firing, creating metabolic strain that shows up as discomfort. Prior research on visual discomfort has identified specific patterns that maximize neural activity: high-contrast stripes, dense dot patterns, and certain color combinations. Earlier studies suggest these stimuli create what researchers call “cortical hyperexcitability,” which may drive up metabolic demand without providing useful information. It’s like forcing your brain to do extra work for no good reason.

The current study extends this line of thinking to complex, real-world images. Pictures requiring more distributed neural activity across visual regions received lower aesthetic ratings. But here’s where it gets interesting.

The Sweet Spot Between Boring and Exhausting

Complexity plays a role, but not in the way you might expect. Extremely simple images don’t rank highest in aesthetic appeal despite requiring minimal processing. A blank white room requires almost no metabolic cost to process, but it also fails to engage the visual system in any meaningful way. The resulting boredom overpowers any benefit from processing efficiency.

This relates to what researcher Daniel Berlyne called the “two-factor model” of aesthetics: preference depends on both avoiding tedium and achieving positive habituation. You need enough going on to keep your brain interested, but not so much that processing becomes expensive. It’s like Goldilocks, not too simple, not too complex, but just right.

When the researchers built mathematical models using activity from each visual processing stage, they found that metabolic costs at all levels contributed to predicting aesthetic ratings. Early visual areas that process basic features like edges and colors played a role, but regions handling higher-level scene and face recognition showed stronger effects. Efficiency matters at every stage, but the metabolic costs of recognizing meaningful content carry more weight than the costs of detecting simple features.

Interestingly, regions in what’s called the brain’s default mode network showed the opposite pattern. These areas, which activate when you’re daydreaming or thinking about yourself, displayed more activity for images people rated as more aesthetically pleasing. This finding aligns with previous research showing that contemplative aesthetic appreciation involves different neural circuits than immediate gut reactions.

What This Means for Understanding Beauty

The results provide a biological foundation for a concept psychologists have studied for decades: processing fluency. When stimuli are easy to perceive and understand, people tend to evaluate them more positively. The current research suggests this fluency has a measurable metabolic component rooted in the efficiency of neural firing patterns. It’s not just that things feel easier to process, they literally cost your brain less energy.

The framework also offers a straightforward explanation for why prototypical examples of categories receive higher ratings than atypical ones. If the visual system has optimized its representations around common patterns (like average faces), processing typical examples requires less metabolic expense than unusual ones. Your brain has essentially created shortcuts for common patterns.

Previous research using deep learning models to predict aesthetic ratings has focused on what features are present in images (like colors, shapes, or composition). The Toronto team’s approach differs by examining the properties of the neural code itself, measuring how the brain actually encodes the information. They found that both the number of active neurons and the total amount of neural activity predicted aesthetic preferences, with total activity providing slightly better predictions.

The Limits of Energy-Based Beauty

The researchers emphasize an important caveat: their findings apply to immediate gut reactions during passive viewing, not to deeper aesthetic contemplation. When people deliberately engage in evaluating art or reflecting on personal meaning, different brain regions become involved and metabolic cost may matter less.

Consider the difference between glancing at a painting as you walk through a museum versus standing in front of it for five minutes, thinking about what it means to you. The first experience might be driven more by processing efficiency. The second involves conscious thought, memory, and emotion in ways that might override or interact with the efficiency signal.

Cultural factors, personal experiences, and artistic training all shape aesthetic judgments in ways the current study didn’t address. The metabolic efficiency signal may establish a baseline preference that other factors can override or modulate. Someone trained in abstract art might find efficiency in patterns that initially seem chaotic to an untrained eye.

The study also used averaged ratings across many observers, which reveals common patterns but hides individual differences. What costs one person’s brain a lot of energy might be easy for another, depending on their visual experiences and expertise.

Despite these limitations, the findings mark the first large-scale demonstration of a direct relationship between visual processing costs and aesthetic pleasure. The convergence of evidence from artificial neural networks and human brain imaging strengthens the case that metabolic efficiency plays a genuine role in shaping visual preferences.

Beauty may not be entirely in the eye of the beholder. Some of it appears to be in the energy budget of the beholder’s visual cortex.

---