Personal computing has a large and growing environmental footprint

It's easy to think otherwise. Our devices are smaller and quieter and cooler than ever before and draw very little power from the wall. But the power computers consume while in operation is trivial compared to the power used to extract and refine the non-renewable raw materials they are built from, to manufacture them and to ship them. The average computer user today has more devices, and replaces them more frequently, than ever before. Often the replacement happens only because the manufacturer has stopped issuing essential security updates for perfectly functional hardware, or because one small part has broken and it's cheaper to replace the whole thing than to fix it, or even for the sake of fashion, because marketing has made the latest smartphone a must-have accessory. E-waste is building up around the world faster than solutions for what to do with it are ever likely to be found. Vast quantities of power and water and raw materials are also consumed behind the scenes where the end-user doesn't see or think about them— in data centres, in mobile phone infrastructure, in undersea cables. The pursuit of ever higher video and audio quality, and the norm of streaming the same content over and over instead of downloading once and storing locally, mean that our appetite for data risks growing faster than the energy costs of sending it can be reduced. That risk is even greater with training modern machine learning tools. Today's “paperless” world isn't green, it's filthy.

Computers and networks are not inherently good

The idealistic visionaries who embraced personal computing and networking early on were very optimistic about the societal consequences of a wired-up humanity. The “global village” was supposed to connect people from all around the world and increase our understanding of one another, reducing conflict. Fast and cheap access to information was supposed to make us all better informed and better educated. Reduced costs and logistical hurdles to publication were supposed to democratise the media landscape and dissolve the old entrenched media monopolies. Autocratic regimes were supposed to crumble as their populations bypassed censored local media. None of these Utopian visions have come to pass. Computers and networks have proven to have the capacity to be addictive, to disrupt attention and to harm mental health, have been used to spread misinformation, to polarise public discourse, to radicalise people and to keep whole populations under tight surveillance. Where old monopolies have crumbled, they've been replaced by newer, even more powerful ones. Computers and networks certainly have done good things too, and we shouldn't change from thinking of them as inherently good to thinking of them as inherently bad, but we can't keep taking seriously the old foundational myths that they are automatically a net benefit wherever they are introduced, that they naturally empower individuals against corporations and governments, or that they inevitably spread the internet's pioneering ideals of decentralisation, decommercialisation and censorship resistance. Those stories just don't hold up anymore.

Better hardware, better software and better networks are possible…

The short useful life of modern devices, the difficulty or prohibitive cost of repair, and the proliferation of e-waste are not natural properties of personal computing, but are things we do to ourselves. Similarly, software and networks do not need to be addictive or harmful, but are intentionally designed to be so in pursuit of profit and power. Recent years have seen a slow but steady increase in awareness of these realities and efforts to counter them. Computing hardware can be designed for easy repair and easy disassembly as higher priorities than easy manufacturing and compact size, using standardised parts available from multiple manufacturers, with all design information shared freely and openly. New software can be written for still functioning devices whose manufacturers have carelessly abandoned them. Simple, highly portable virtual machine designs can abstract over the incompatibilities of diverse hardware, providing a stable common denominator so that the lifespan of software and digital media are not tightly coupled to the availability of specific machines. Network protocols for sharing files without tracking users or easy insertion of advertising are both easier to implement and more energy efficient than protocols which include those misfeatures. These efforts should be celebrated and encouraged and emulated, but we should recognise them as partial solutions, not whole ones.

…but computing less is possible, too

Designing and building greener hardware and software and more humane online spaces is the easy and immediate reaction of green-minded and socially conscientious hackers to some of the realisations above, because it's fun and challenging and creative and rewarding in all the ways that we love. But we should be questioning and changing not just how we compute, but when we compute and why we compute. Green urbanists know that making cars safer and more fuel efficient is well and good, but improving public transport and infrastructure for cyclists and pedestrians so that fewer cars are needed is even better. Green hackers should think along similar lines. The environmental impact of computing can and should be reduced, but it can never be eliminated. Even if it could, ubiquitous computing combined with always-on global networking can clearly create problems just as easily as it can solve them. Often the problems solved are relatively trivial, while those created are serious and only become obvious in hindsight. Treating computers and networks as universal solutions to every problem is “smart but not wise”, and we who enjoy using them must work hard to scrupulously avoid this trap. After all, a time when most people's daily lives did not directly involve computing at all is still within living memory. Life might be more convenient today, but it would be very hard to argue that most people are happier, healthier and more fulfilled now that they're constantly online.

Keeping perspective, finding purpose

We don't need to get carried away. The environmental impact of computing is real and rising, but is currently small relative to that of cars, beef, single-use plastic, mass tourism, or fast fashion. We need to compute less, but less doesn't have to mean not at all, especially when we are surrounded by functioning computing devices which have already been made and therefore already inflicted most of the ecological damage they ever will. Arguably, we're morally obligated to get as much use out of them as possible, so the terrible price paid for them was not in vain. This is good news for green souls who are also incurable “computer people”, because figuring out how to keep those devices running for as long as possible with as little power as possible will provide a lifetime of interesting technical challenges. But the technical challenges are the easy part. We will have to contend with cognitive dissonance and spiritual emptiness while tinkering away if we don't have a convincing answer to just why we're so invested in keeping personal computing alive. Are we just trying to cling to our favourite toys for as long as we can, with as little eco-guilt as possible? Or do we actually believe that low-impact computing devices can do enough ecological or social good to be worthwhile? If so, are we just repeating our history of uncritical Utopian thinking about computers? These are more important questions for a “sustainable computing” movement than any about hardware or software!

Can we fight fire with fire?

The environmental harm from personal computing is a small part of a much larger crisis. That crisis, at its core, is not a technical problem, requiring a technical solution. Rather, it is a cultural problem, driven by holding up endless progress as our species' foundational ideology and by the perception of the natural world as something we exist outside of and, thanks to our brilliance, can exploit endlessly without consequence. Cultural problems require cultural, not technical, solutions. Nevertheless, greener personal computers and networks may still be productive tools for the job. This is no paradox! We tend to think of computers as dispassionate number-crunching, problem-solving machines, but much of today's personal computing is only “incidental computing”. After all, the daily lives of well-adjusted people rarely involve non-artificial problems which warrant high-speed automated data processing. Personal computers are conquering the world, and its landfills, mostly as smaller, faster, more portable and flexible evolutions of non-computational tools like the telephone, typewriter, camera, radio, and television. Like their analogue predecessors, our digital devices and the global networks they form are fundamentally cultural artefacts, ones with unprecedented power to transmit ideas, attitudes, and beliefs, to shape worldviews, and to influence even our offline habits and behaviours. This cultural powertool has so far only exacerbated the larger crisis, by perpetuating the ideals of infinite technological progress and increasing separation from the natural world (even disdain for it, expressed with the hacker's pejorative “meatspace”). Could it instead be used to reverse this indoctrination? To replace it?

Deep green memetics

A green computing movement might just make sense if the focus is far less on the devices and technologies themselves than on what we do with them and, most importantly, on how we are changed by doing those things in the long term. We need to figure out how people might use long-lasting, low-energy computers to do genuinely interesting, fun and engaging things which have the side effect of challenging and changing our ideas, beliefs, behaviours and habits in ways which strike at the sociocultural roots of environmental crises, and which make a future with fewer devices and less screentime seem rich and rewarding instead of unbearably dull. Can computers and networks redirect our attention toward the natural world instead of away from it? Can they help us to better understand that world, to feel connected to and responsible for it, encouraging us to make changes in those parts of our lives which have more environmental impact than computing? Can they make us genuinely want to spend less time in front of screens and more time outside? Can they facilitate long term thinking instead of obsession with novelty? Can these changes be initiated with computers but persist without them? The only way to find out is to try! Let's hack software and hardware not for their own sake, but in order to hack hearts and minds, to convince ourselves the latter can still flourish in a future with less of the former, and to help bring that future about.

A call to arms

Let's hack our modems and routers to throttle our internet connections based on local weather conditions, so they're slowest when it's nice outside and fastest when it's dark and wet. Let's write blogs about our outdoor hobbies and activities. Let's write software to identify wildflowers and fungus. Let's use computers to design beautiful and/or useful things which can be made by hand from natural materials. Let's create video games which teach basic ecological principles. Let's use abandoned smartphones running old, insecure operating systems in airplane mode to take time-lapse photographs of sunsets or field recordings in forests. Let's use those photographs and recordings as raw material for computationally efficient procedural generation engines to create nature-inspired digital artwork that fits in as few bytes as possible. Let's create and share free resources and ethical networks for nature-focussed autodidacticism, citizen science and citizen journalism. Let's use GPS-enabled devices not to access road-centric online map services but to create our own nature-centric maps of our local areas, so as to learn their geography and geology and hydrology. Let's apply biomimicry to software. Let's launch low-bitrate streaming radio stations which synthesise music in real time based on environmental sensor readings, so local communities can listen to their sunrises and storms together, and actually hear their local climate changing over the years. Let's do all of this and more, treading as lightly as possible as we go, and let's freely share everything we learn and just as importantly everything we feel while doing it.