We are not designing for users. We are designing for Paleolithic hominids.
Every roadmap talks about "the user", a rational, task-focused person who reads our screens and completes our flows. That person doesn't exist. The one on the other side of the glass is a primate with a nervous system shaped long before screens, now asked to process notifications, menus and dense type all day.
This article looks at a few places where biology and interface design meet, and where the evidence is solid enough to design around: the energy cost of thinking, how people actually scan pages, why touch feedback matters and what kind of spaces feel safe. It's for designers who want principles that last longer than the current interface trend.
§ 01
Thinking is expensive
The brain is about 2% of body weight and uses about 20% of the body's oxygen, and therefore its calories. That's the classic estimate from Raichle and Gusnard in Appraising the brain's energy budget, and it explains a lot about how people behave in front of a screen. For most of human history, spare energy was scarce. A brain that spent it freely on puzzles that didn't matter was a liability.
I find that a useful frame for cognitive load. It's tempting to say a cluttered dashboard registers as a physical threat. The evidence doesn't go that far. What it does support is simpler. People conserve effort by default. Faced with a screen that demands work before it gives anything back, many of them leave, skim or guess.
In product terms, every decision you push onto someone has a cost, and they pay it from a small budget. Defaults, sensible presets and fewer choices per screen aren't minimalism for its own sake. They're how you stop spending a person's attention on things they didn't come to do.
§ 02
People forage pages, they don't read them
Nielsen Norman Group first described the F-shaped scanning pattern in 2006. People read across the top, then a shorter line further down, then run their eyes down the left edge. Later research added other patterns. Some people scan only headings, some hunt for a specific word or number and a few highly motivated readers take in everything.
The F-pattern isn't a law to design toward. It's what happens when a page gives the eye nothing better to follow. That's the useful insight. Scanning is foraging. People sample a page for signs that the thing they want is nearby, and they only commit attention when the signs look good.
So give the forager good signs. Put the most important point in the first two lines. Write headings that start with the words people are looking for. Put numbers where they'll be spotted. And keep motion away from the edges of the screen when you want someone to focus, because movement in peripheral vision pulls attention whether we like it or not.
§ 03
Glass gives the hand nothing back
For almost all of history, tools pushed back. A key had travel, a lever had resistance and a bowstring had tension. A touchscreen is flat glass that feels the same whether your press worked or not.
That gap has a measurable cost. In 2008, researchers at the University of Glasgow added small vibrations to a touchscreen keyboard so each key felt like it had edges and snapped back. They found that without touch feedback people made many more errors, and that tactile feedback brought touchscreen typing close to a physical keyboard. Researcher Eve Hoggan told the university that feedback through the finger is more natural and the brain can process it faster than visual feedback.
That's why micro-interactions aren't decoration. A button that visibly depresses, a toggle that snaps and a light haptic tick on a successful action all tell the body that its command landed. Without them, people check the screen again, tap twice and lose trust in the interface without being able to say why.
§ 04
Engagement that leaves people worse off
Feeds and notifications borrowed a trick from slot machines. Rewards that arrive on an unpredictable schedule are hard to stop checking. It works on the metrics dashboard. Over months, it tends to produce users who feel worse and leave as soon as something less draining appears.
Humans did evolve to care deeply about status and belonging, and products can serve that need without exploiting it. Shared rituals, clear norms and the feeling of being known by a small group keep people coming back for good reasons. A weekly summary instead of a constant drip, a feed that ends, notifications batched into a few predictable moments and a clear way to step away all respect a person's attention instead of mining it.
This is also a business argument. Products people trust get recommended. Products people resent get uninstalled the day a competitor shows up.
§ 05
Safe spaces have a view and a way back
The geographer Jay Appleton proposed prospect-refuge theory in the 1970s. People feel at ease where they can see their surroundings and also have somewhere sheltered to retreat to. Think of a cave mouth overlooking a valley, or the corner table in a restaurant.
Interfaces can offer both. Prospect is clear navigation, breadcrumbs and a visible sense of what's ahead in a flow. Refuge is undo, drafts that save themselves, a clear way home and privacy settings that are easy to find. People explore more boldly when they know they can get back.
Visual texture matters too, within limits. Physicist Richard Taylor's research on fractal fluency argues that human vision processes the mid-complexity patterns of nature, like trees and clouds, with particular ease, and links them to lower stress. It's an interesting line of research for imagery and backgrounds, though most of it comes from physical environments, so treat it as a direction to test rather than a rule.
§ 06
Design for the animal, test with the person
None of this replaces research with real people. Biology gives you good starting assumptions. Your users tell you which ones hold for your product.
If I had to compress it into a checklist, it would be this. Spend people's attention as if it costs them something, because it does. Put the signal where a scanning eye will find it. Confirm every action through more than one sense. Don't build rewards that work like slot machines. Give every flow both a view ahead and a way back.
Interfaces will keep moving, from screens to voice to AI agents to whatever comes next. The nervous system using them won't change on any timescale that matters to a product team. That makes it the most stable design constraint we have.