evolutionary aspects of interface design

Theory of evolution prescribes rules for how spontaneous structures emerge. In other words, at its roots, it is a theory about design. Hence its reach extends far beyond biology.

As designers, we tend to think that our products are results of our own independent design choices. But, in reality, our decisions are largely governed by macro factors and creativity is mostly a cultural phenomenon.

There is no need to get pissed off about this though. Instead, we should study how evolution exhibits itself in interface design, so that as designers, we can better understand the contours of our own individualities.

Fitness

Unfit species are doomed to perish by natural selection. Similarly, users make sure that bad designs get punished with lower adoption, conversion and engagement rates. After all, you can not force bad design onto users.

Note that the product design as a whole faces natural selection pressures. UX is just part of the story. Many products with incredible user experiences have disappeared due to entirely business related reasons.

This creates an argument for setting up “genetic repositories” like the Little Big Details blog. There is a need to preserve good UX specimens because their container products sooner or later slip down the fitness landscape.

Randomisation

The fitness landscape is laden with local maximums which are like small hills whose smallness can only be recognised from a bird-eye view. Evolutionary processes climb local maximums via iterations and avoid being locked into one via randomisations. In other words, the global maximum can not be reached solely by iterative methods.

Similarly, A/B testing small differences one at a time will not get your design any nearer to the global maximum. You need to be bolder than that.

Affordance

In biology, form and function are inextricably entangled with each other. It is impossible to make a general statement about whether form dictates or follows function. For instance, feathers were initially “invented” for heat preservation, but later mediated the flight of birds.

Similarly, the notion of affordances in interface design ascribes fluid functionality to forms. Good designers know that there is a big difference between intended function and actual function. That is why there is always an element of surprise in user testing.

Convergence

Given a constant environment, there is always an optimum way to achieve a certain task. And evolution somehow manages to find a way to converge onto this optimum, even from wildly different genetic lineages. For instance, squids and mammals both have camera eyes although their last common ancestor was probably a blind creature.

We see convergences over time in interface design as well. For instance, by learning from each other’s mistakes, we finally seem to have agreed on what an optimal check-out experience should look like. Of course, these kinds of convergences eventually result in a total commoditisation which manifests itself in the form of themes and templates.

One interesting convergence we see at the moment is taking place in the social media sphere. I don’t know if you have noticed, but all social media platforms have started to look like each other. Just consider how Twitter’s post structure has evolved over time! Soon the only thing that distinguishes one platform from another will be its community of users. (This, by the way, happens to be the same equilibrium that online forums have settled into years ago.)

Mimesis

Genes can be transferred either vertically via sexual and asexual reproduction mechanisms or horizontally via other means. For instance, bacterias can borrow genes from each other to defend themselves against a common enemy. (This is how anti-bacterial resistance develops.)

Similarly, good design elements spread really fast. Social platforms like Dribbble and curations sites like Awwwards can be seen as horizontal transfer mechanisms. Designers copy each other intentionally or get inspired from each other all the time.

Mutation

Of course, even copying requires a certain level of base skills that most mediocre designers lack. These sorts of imperfections introduce mutations into the horizontal gene transfer processes. Bad mutations often build up overtime and result in horrible copycats.

Rarely, good designers improve on the work of other good designers and introduce good mutations. These sorts of positive feedback mechanisms become the main drivers behind the formation of new trends.

In either case, mutations introduce variations just as they do in genetics.

Interdependency

Evolution works on different scales and on each scale it introduces interdependencies between the participating entities. For instance, survival of a single organism depends not only on a vast ecological network outside, but also on an equally complicated protein network inside. Each of these networks contain interdependencies which can easily go haywire from outside shocks.

Thorny relationships exist in interface design as well. A UX problem feels challenging precisely when all the solutions you can come up with seem to introduce other UX problems. Faced with such a situation, you have no choice but to engage in a balancing act. In other words, you need to step back and consider the big picture.

There is no absolute perfection in design. There are always some trade-offs. Bad designers are just not aware of them. The challenge is to find that sweet spot that is close enough to the global optimum.

Modularity

Modules are hierarchically nested, specialised building blocks. They are ubiquitous in nature. For instance, your body has organs, your brain has lobes, your lobes have neurons, your neurons have organelles and so on…

The leading hypothesis is that modularity mainly emerges because of rapidly changing environments that have common subproblems, but different overall problems... Intuitively, modular systems seem more adaptable, a lesson well known to human engineers, because it is easier to rewire a modular network with functional subunits than an entangled, monolithic network.

- Evolutionary Origins of Modularity

Hence, modularity is very fundamental in the sense that it is both a product of evolution and a lubricant increasing evolvability.

Modularity is a deep feature of interface design as well. However, as in evolution, it often takes a substantial amount of time for it to surface:

Designers must have significant knowledge of the inner workings of a system and its environment to decompose the system into modules, and then make those modules function together as a whole. Consequently, most modular systems that exist today did not begin that way — they have been incrementally transformed to be more modular as knowledge of the system increased.

- Universal Principles of Design

Once a design becomes modular, it becomes more accessible to tinkerers. Professionals focus on multiplying and improving the modules themselves, while amateurs spark an explosion of creative combinatory activity at the higher level. (Think of website building tools like Squarespace.) Again, as in evolution, modularity emerges both as an end-point and a starting-point of new design processes.

However an important difference remains between natural and man-made modularities. It is best framed as the difference between what is complex and what is complicated:

An engineered system is complicated, while an evolutionary system is complex. In the case of complicated systems, the pieces can be disassembled and reassembled again, and the function of the whole can be guessed quite well from the functions of the parts. In the case of complex systems, the function of the whole is an emergent property of the parts, and in most cases we cannot make a straightforward guess about the function of the top network if we only know the function of the bottom networks (modules) in a piecewise manner.

Peter Csermely - Weak Links (Page 98)

The stickiness between the modules in the case of evolutionary systems stems from the necessity that biological structures need to be robust since they are under constant noise bombardment.

Plasticity

Plasticity is an organism’s ability to change itself within its own lifespan. It is the evolution’s way of encoding an anticipation of reasonable amount of environmental variance in advance. Something like controlled, fast-paced, micro-scale “evolutions” riding the pseudo-random, glacially-slow, giant waves of evolution!

Our muscles respond to more work-outs, our neurons respond to more challenging exercises, our guts respond to more food in-take etc. There are countless examples.

Plasticity in interface design manifests itself as responsiveness and personalisation. Both are currently at a very primitive stage. Even building a website that redesigns itself with respect to different screen sizes is quite difficult to do well. Facebook customises your feed and Google personalises your search results with respect to the data they gather about you from various sources. But gathering data and understanding it are two completely different challanges. We have yet to see an impressive example of interface plasticity. I guess Data Science is still at a very early stage of development.

Vestigiality

Many organisms carry around vestigial organs that used to be highly functional back in their respective genetic lineages. Evolution seems to be inherently less capable of simplification than enrichment. A substantial chunk of our DNA is claimed to be junk. Just meditate on this a little!

Similarly, over time, interface designs accumulate vestigial elements which need to be cleaned up periodically. For instance, a user flow path that was popular can all of a sudden be abandoned upon the introduction of a new screen or a shortcut.