⨳ 19 Nov 2021
Everything digital costs something physical. It’s time for a digital conservation movement.
The digital world is a parasite that we hypnotically spoon-feed like a baby from what we assume is an inexhaustible puree of attention, time, content, energy, and resources. That we don’t see it this way — that we see our relationship to the digital world in almost exactly inverse terms — has created a profound civilizational threat that masquerades as an opportunity.
The truth is that there is no such thing as the “digital world.” It is not a realm that exists apart from the so-called real world. Everything that is digital — information, exchanges, and experiences — is also physical. And yet, as infinite as we imagine the digital world to be, there is only so much physical stuff to go around. We are facing a very real future of scarcity, not abundance, and it will only be more severe if we continue to deny the true, hard costs of digital culture. We need to begin to express it in harmony with the physical world.
The sad irony of the technological culture spreading over this planet is that its forward momentum seems entirely driven by a desire to escape it. Some of the largest corporations in human history are staking their immediate future on building an inhabitable digital world in which they can claim our every interaction. Others are spending their digital profits on finding ways to blast off this planet and colonize new ones. Meanwhile, the sub-current of longer-term technological thinking is flowing on the desire for and expectation of a singularity — of a point at which technological growth enters a “runaway reaction” of autonomous self-improvement — that will, in theory, create technology capable of housing our consciousness and cheating death. The cost, again, is living within a machine for all eternity. Is something so wrong with this real world that we’ll do anything but invest in it? I wonder why we’ve given up on living here — on making this place our home, rather than a projection, a rocket, or a server farm.
As critical as I am of each of these ideas, I also recognize that they are not actually technological in nature. They are technological choices born of a particular worldview. Changing them is not a matter of changing technology, but of changing minds. And if we’ve learned anything from recent history, minds aren’t changed by cogent arguments, or by proof, or by discoveries and innovation, but by repeated, simple sloganeering. Minds are molded slowly and subtly, not remade by singular applications of logic.
One way our minds have been shaped is by the way digital experiences look and feel. The practical irony of technological culture is that the more resource-hungry it becomes, they less it refers to the world on which it depends. The visual language of the digital world is increasingly anodyne and sterile, purged of elements that feel human, organic, or personal. It’s a neutral aesthetic that expresses the point of view of machines looking upon people, not the other way around. Our faces are encircled avatars, and our spaces are seen through rectangular viewfinders. We obsessively feed machines with our images and observe the world through their digital pastiche. And so far, our experience has been primarily visual. We see all this stuff and we say we feel it, but not in a way having anything to do with touch. The digital world is still just a projection.
Within the projected digital world, we’ve experimented with many aesthetic forms. Early digital spaces were crude and simple. As we gained processing and rendering power, they verged toward the skeuomorphic, representing physical surfaces in pixels. When we realized how kitsch that was, we veered in the opposite direction with “flat design,” a reactive two-dimensionality that we are slowly letting go of in favor of something a bit more spatial but still without interior reference to the organic. It remains a synthetic aesthetic which sees the organic as though through a window, looking upon it but not existing within it.
The more time we spend in this projection, the greater our impression of it as a self-sustaining, independent reality. One that exists on its own terms, not as dependent of the ground beneath our feet. We need to adapt our digital visual languages and interaction patterns to make frequent, intentional references to the physical world in order to build a better understanding of where it is and where we are when we inhabit it.
That the digital world is dependent on the physical world is self-evident. But as I’ve explained so far, it is too easy to remain in ignorance or denial of that due to how the digital world works and looks. Language also plays a role here, as we’ve often chosen metaphors which accentuate the non-physical aspects of digital information and downplay their draw upon tangible resources. “The Cloud,” for example, describes a place in which digital information is stored as if it is a vapor — a light, airy, and ephemeral thing — rather than an enormous collection of fuel-dependent machines that alter our cloud-strewn atmosphere as much as all the oil-burning jet engines that traverse it.
“The Cloud” is actually a collection of many data centers. Data Centers are actually enormous buildings full of networked computers. They require the same resources that any building requires in order to be constructed and maintained, but exponentially more; data centers require 10x to 50x the energy per square foot of a typical office building. In the United States, they are responsible for 1.8% of all electricity usage. This may seem like a small portion, but when you consider that the population of the United States is 330 million people living in 140 million residences and working in 6 million commercial buildings totalling almost 100 billion square feet, suddenly 1.8% is a substantial portion. The Department of Energy considers any server cluster a data center, even the couple in a random business’ networking closet, and so they estimate the total number of data centers in the United States at over 3 million. But when you define a data center as a commercial entity and structure devoted to the centralized storage and transmission of digital information, there are only 2,670 of those. So, .002% of the structures in this country account for almost 2% of its electrical need. Wow.
However, even the most basic research into data centers reveals that many of them are far more efficient than a typical structure. Their efficiency is measured in terms of Power Usage Effectiveness (PUE), or the ratio of the total amount of electricity consumed to the amount of electricity delivered to its equipment. The closer PUE is to 1, the more efficient the data center is. The largest data centers in the United States have published PUE of less than 2; the industry average is 1.5. Facebook, my personal technological sin eater and absorber of digital angst, has an average PUE of 1.1. That’s pretty impressive, I must admit.
Data centers achieve low PUE by operating on larger shares of renewable energy, like solar and wind. Facebook’s low PUE is even more impressive when you consider that it maintains 18 data center campuses across the world, which occupy a total of 40 million square feet, all run solely on renewable energy. Many data centers operated by the most visible digital companies are similar. Google, Amazon, Microsoft, and Netflix all invest in renewable energy to fuel their data centers, but nowhere near 100% of the time. Most are still depending upon fossil fuels some of the time. Meanwhile, the majority of data centers in the United States are run by companies you probably haven’t heard of. The six largest, publicly-traded data center operators in the United States are Equinix, Digital Realty, CyrusOne, CoreSite, QTS Realty Trust, and Switch Inc. Equinix alone maintains over 85 individual data center campuses. By size, the top ten largest data centers in the United States are mostly run by these relatively unknown companies and serve private industry and government, not consumers, except for two: Facebook, sixth on the list, and Microsoft, ninth. Ranked fifth is the NSA’s largest data center known as Bumblehive.
Electricity isn’t the only resource that data centers consume out of proportion with the rest of our inhabited structures. The more machines that are clustered together, the more heat they generate. But a hot environment isn’t good for computing, so data centers need to be cooled to keep their machines running optimally and from overheating. Most data centers do this by evaporating water to produce cool air. Together, they use almost 30 billion gallons of water per year to cool their machines. Mitigations of this consumption, like reducing potable water use, on-site water treatment, rainwater harvesting, and natural water-source restoration, are increasing but all require additional energy to perform.
According to Nature, data centers “use an estimated 200 terawatt hours each year. That is more than the national energy consumption of some countries, including Iran, but half of the electricity used for transport worldwide, and just 1% of global electricity demand.” The words “half” and “just” are doing a lot of work there. That the combined energy consumption of data centers is half of all transportation vehicles throughout the entire world,” and just 1% of all the electricity being used by the entire world is staggering. This is the cost of digital culture.
When we widen our scope to the entire information and communications technology ecosystem, including all our individual devices, not just these unseen stores of our data, the situation is more extreme. All of it put together accounts for more than 2% of global emissions. And before you let out a sigh of relief that it’s only 2%, remember that all global emissions include every emission of every vehicle, chimney, fire, and cow’s ass across the entire planet.
Today, the global “datasphere” is measured at over 33 zettabytes of stored information and is projected to increase to 175 zettabytes in just five years. (A zettabyte is 1,000 exabytes; an exabyte is 1,000 petabytes; a petabyte is 1,000 terabytes.) What is it all? Where does it all come from? Well, we sent 306 billion emails every day last year, for starters. We uploaded over 350 million photos every day to Facebook alone. We Tweeted over 500 million times every day. I’m not sure we can even comprehend these numbers. But we can understand that they are a lot and that they are why the number of data centers is increasing and why resource-consumption and depletion is exponential.
We should give some thought to the share of these numbers that is ours, personally. How many emails do you send? How often do you share photos, or Tweet, or search with Google? You don’t need to know the actual number. But consider that each Google search you make releases 5-7 grams of carbon dioxide (CO2) into the atmosphere. Every email you send releases 20 grams of CO2. If you send 30 emails per day for a year — a modest number by some peoples’ standards — you’ve released as much CO2 as if you’d driven from Boston to Philadelphia and back (a bit over 600 miles).
Considering these numbers should alter your environmental awareness. Thinking twice about commonly thoughtless email behavior — forwarding, replying-all, archiving everything — and Google use — typing URLs directly rather than searching for them — may actually be as ecologically virtuous and impactful as recycling an aluminum can. Tiny ripples in a sea of digital activity.
A few years from now, the “digital world” will cost us trillions per year and pollute the physical world more than any one country, more than all the travelers moving about its every square inch.
More than half a century ago, Buckminster Fuller critiqued our dependence upon fossil fuels in light of the abundance of renewable sources we were so fortunate to have on this planet. He popularized the idea of Spaceship Earth, a phrase meant to remind us of our mutual dependence as we move forward toward the future. For Fuller, it was important to be aware — at all times — of one’s actual place in the world, as a person within a system, within an economy, within a society, within nature. Reality, rather than being a conglomeration of smaller, freely floating bubbles of ideas and things and groups that periodically bump into one another, is instead a nested structure. Similarly, our “digital world” is nothing of the sort. It is not a world, as The World is, nor is it even a primary sub-set of The World. The digital experiences we create are a subset of the machines that we create, which are a subset of the natural world, which is itself a subset of all information. In this structure, the natural and build world are partners in filling the so-called “digital world” with the basic building blocks of reality.
When Bruce Mau framed sustainability as “the welfare of all life as a practical objective,” he properly situated our actions in this nested reality. There is no digital world without the natural one. There is no natural world without its inhabitants, humans and every other living thing. And there is no life without a natural world to support it. Sustainability, ecology, conservation — whatever you want to call it — preserves this cyclical balance. It is what we must do when we can no longer claim ignorance. The next few decades will be about that kind of action. It will be about balancing the digital and the organic; not necessarily about shrinking the growth of the digital world, but about calibrating our experiences within it.
How can we be more mindful? Imagine if you had to pour out a gallon of water every time you Googled something. Or if you received a carbon “bill” for every email you sent or received? Or if you had to store up sunlight before you could watch a YouTube video? We really shouldn’t have to imagine these things. They are happening. But we lack the conscientiousness to connect our digital consumption to its very real depletion of natural resources. Individually, we bear that responsibility. But in addition to corporations feeling a social pressure to power their operations with renewable energy, I would also like to see them publish realtime energy “balance sheets” on their screens. Imagine Google.com with a realtime tracker of energy and resource use along side its search results. Imagine a small solar icon with a readout of the hours it takes to generate the wattage needed every time a search results page loaded. Imagine an animation depicting a gallon of water slowly pouring out as you scroll down the page. We need these cues to directly connect the digital to the physical. Without them, that connection is hidden away in anonymous warehouses that few of us will ever see, burning resources at a scale we’ll never understand.
Everything digital is also physical.
Does digital matter exist? Or is that an oxymoron — akin to asking about hot ice or square circles? Taken at face value, digital matter may seem like a contradiction in terms, but it is actually a necessary truth that we must reconcile with. Because while we continue to build larger, denser, and more complicated digital “worlds,” the real world — the planet on which you and I live and every single thing we make depends — is becoming a less and less sustainable support for our existence.
It wouldn’t be fair to chalk it all up to a semantic misunderstanding; reality may be understood in the mind, but it is experienced under foot. And that, right there, is the root of the problem. The conceptual and the concrete are coeval in reality, and yet if there is a defining concept of the modern era, it is probably the schism between the two. Were it not possible to separate reality in this way, I don’t think that the largest corporations in history would have been built upon digital worlds. But they have — all of them.
And while the take-wars heat up over the Metaverse and its competing visions for a more immersive and breakaway digital civilization, we still have time to consider whether we will join them or not. You have many reasons not to; let ecology be another. Let it rise to become a primary reason, because you now know that attention is not just another way of expressing time. It is a natural resource. It is part of you as an organism in a living, breathing ecosystem made more fragile every moment by every byte.
How much disdain for actual reality must you have in order to want to spend even five minutes as a legless, bloated Sims character in Mark Zuckerberg’s cartoon spyhouse? I don’t know. I’d say some. But how much disdain for actual reality must you have to build such a world in the first place? That I know, and it is a lot. This world — the one in which smaller, digital “worlds” will come and go, on which they will depend — is your base reality. It is the only thing that matters. There is likely no future without digital experiences; I wouldn’t argue for that anyway. But there is certainly no future without a proper understanding — at the individual human level — of the costs of digital experiences. Every bit costs something. Now, how much are you willing to pay?
I needed this primer on Web3. I also needed Robin Sloan’s take for the cautiously curious. How to start an attention diet. I love everything about this — Why Do We Interface, by Ehsan Noursalehi: the design of it, the thinking within it, the classification of it as a “micro-book with incomplete observations.” These photographs turn reality into a riddle. An oldie but a goodie, Listen to Wikipedia.