Increasing improvements in user interface design are a natural result of the physical laws of the universe. To understand the underlying meaning of that statement, and its potential consequences, it is necessary to enumerate a sequence of logical assumptions (which I will attempt to summarize, but are no replacement for a fuller reading of the sources cited below):
1. The universe is driven by physical forces, including entropy – the dispersal of energy from localized concentrations towards generalized equilibrium. The cliché example of entropy is the spread of heat from a cup of coffee to the surrounding room. The heat spreads from the local area (the coffee cup) to dissipate equally within the room.
2. The origination and evolution of all systems in the universe from stars, to rocks and eventually life, has been driven by entropy. Moreover, over the history of the universe systems have developed that have been progressively more effective at increasing entropy (i.e. distributing energy). Astrophysicist Eric Chaisson has applied the measure of energy rate density – the flow of energy per unit timer per unit mass – as a common metric that can be used to compare the relative energy emissions of everything from stars, to plants, to people, to cities. His chart of energy rate densities from an upcoming publication, illustrates the emergence of higher energy density systems over time. While stars are incredibly powerful, they are also incredibly massive and pound-for-pound do not emit the relative energy of plants and animals (incidentally, “Gya” is shorthand for giga years, as in billions of years ago).
3. A key contributor to the entropy contributions of living organisms is the complexity of their behaviors – for example, digestion, mobility and cognition. Across species, mammals have higher energy rate densities than reptiles, which are higher than fish, and so on. The more highly evolved the animal, the higher the energy gradient (birds being the interesting anomaly to this pattern due to the high energy demands of executing flight in three dimensional space.)
4. Technology originated and evolved with life, to the point where man-made technologies have even higher energy rate densities than the humans that created them. Kevin Kelly illustrates this (using data drawn from Chaisson), in his recent book, What Technology Wants. As Kelley illustrates below, computer chips, with their extremely high ratio of heat output to mass, have energy gradients that are among the highest of any system in the history of the universe. In other words, the evolution of systems has followed a path of progressively higher entropy output, starting from large scale inanimate systems (stars, planets) to smaller scale living systems (plants, people) and presently to even smaller scale technological systems.
Interfaces as Entropy Enablers
If you’re still with me, you may be thinking that this is all interesting, but how does it relate to user interface design? We can consider the implications at a couple of levels.
Superficially, user interfaces are the mechanisms by which we control computer chips, how we interact with these powerful entropy engines. At the least, this should lead one to appreciate that we’re dealing with powerful stuff even if we have grown accustomed to them because of their ubiquity. Chaisson even suggests that more transistors may have been produced than any other product in cumulative human history. From a historical perspective, the activity of designing interfaces which control interactions with these chips could be considered as fundamental a human activity as laying the bricks to build cities.
At a deeper level, we can consider the computer chip as a product and force of evolution and entropy. Like humans (and from humans), chips evolved as a progressively powerful entropy-generating system. But from an evolutionary context, why do chips appear to be a successful system? In evolution many systems or groups don’t not successfully proliferate – for examples the Neanderthals did not spread out or survive very long, compared to Homo Sapiens. In the case of humans, spoken language led to the rapid advancement and growth of the human population 50,000 years ago. Similarly, I posit that graphical user interfaces have accelerated the propagation of the computer chip. Early on, interacting with computer chips was difficult, requiring specialized knowledge and skills. Over recent decades, the refinement and proliferation of the graphical user interface allowed an explosion in the number of people who could interact with computer chips across a range of contexts, activities and platforms.
In other words, user interfaces have enabled a rapid increase in the number and uses of computer chips, directly impacting the rate of energy dispersal. There is a logical and perhaps inevitable sequence to this as the graphical user interface connected the language of the humans to the language of the computer. Taking this a step further, we can consider graphical user interfaces as an adaptive trait of computer systems that led to their successful proliferation, with humans as the carriers.
The Future of Interfaces
User interfaces increase entropy. To an interface designer, this might be taken as encouraging, discouraging, or both. To be discouraged is pointless as entropy is fundamental and inevitable. I see it as encouraging that those aspects of interface design that we should strive for as a profession - accessibility, ease of use, effectiveness, and efficiency – are characteristics that will continue to lead to the evolution and expansion of technologies.
What does this mean for the future of interface design? I would expect a continued expansion of chips with greater power, smaller scale and concomitant higher energy gradients. Similarly I expect interfaces that allow us to keep up with this growth in technology by effectively controlling more power and complexity. At some point in the future, we will reach a limit in what can be done by a user interface just as we will reach a physical limit in what can be done by a computer chip (i.e. the flattening of the Moore’s Law curve). What is the ultimate end state of the user interface and what can apply from entropy in the short-term to influence the design of interfaces? I will consider these possibilities in future discussions.