Ecological Computing

I was born in 1984, which afforded me a front row seat to the insinuation of computers into every aspect of daily life. What's more, my mom was a technology integrator for local schools, so some of my earliest childhood memories involve the clunky sounds and musty smells of early personal computing. I spent a lot of time on the computer, but also outside, exploring the sandy hills and canyons around my suburban Southern California home. It wasn't until many years later that I understood there to be a relationship between these domains.

Every few years, computers became more sophisticated and ubiquitous, but also more opaque. The vast majority of people, never wondering much about how they worked, simply accepted that they did (and became quite annoyed when they didn't). Kids at the time saw their multi-purpose rooms repopulated with new machines every couple years. Meanwhile, somewhere far away and out of sight, a landfill received another pile of the old models, establishing the category of garbage we now call e-waste, the world's fastest growing solid waste-stream.

The manufacture and use of computers requires vast, unsustainable infrastructures. Assembled at the end of extractive and energy intensive processes, the machines and their components are made in huge volumes and then quickly replaced, either chewed up by use or outmoded and replaced with the next year's model. They are made, after all, from rare earths and plastics. They also require a lot of power.

Today, the environmental costs of computation are reaching a new apex with the mass deployment of polluting data centers, all for the sake of feeding continuous positive pressure into the AI bubble. The immense resource demands of at-scale AI development are justified by the promised benefits to humanity and our planet, but of course the real driver is to sustain the doomed fantasy of eternal economic growth.

What if computation could be bent to a different set of incentives? Not the linear, extractive tendencies of capital accumulation, but something more circular and regenerative; something like permaculture? That is essentially the aim of permacomputing, a concept, design principle, and nascent community of practice that seeks to reframe how humans think of and relate to computers. Importantly to me, the principle could also extend to all kinds of other technologies.

Permaculture, put very simply, aims to sustainably meet human needs by leveraging the existing processes of nature, in alignment with the energetic limitations and regenerative cycles of the Earth and its ecologies. Since computers are every bit as much an Earthly phenomenon as ourselves and everything else we make or interact with, it stands to reason that they should also function ecologically if we want them to play a role in a decent future. That's easy enough to say; achieving it is much harder.

The computer industry, like so many others, is fundamentally linear — materials are extracted, processed, and deployed in vast quantities, with little to no reuse of existing chips, enclosures, or any other material. New models are constantly developed, pushing the old ones into secondary markets and, inevitably, landfills or the environment, with very low recycling rates. Each new generation of devices becomes increasingly complex and immune to user repair. Efforts are made to counter this trend, but overall the industry is oriented toward disposability.

Rather than a waste stream, permacomputing frames existing computational resources as something to be extended, reused, and ultimately replaced. In an echo of appropriate technology, the goal is to minimize the amount of energy used and material wasted while also maximizing the amount of social or practical value. In many cases, the 'most permacomputing thing' may be to not use a computer at all, unless absolutely necessary. This may seem a sacrilege to consider—social and civilizational processes grow more complex each day, and increasingly rely upon computers to function. This is true; maybe it shouldn't be, at least not to the degree that we have been conditioned to assume.

"If we stop forcing computers down people's throats, and we find that we don't need computers anymore. There's definitely a case where a lot of people would be happier, their lives would be simpler, if computers were just not there and if they were not forced to use them. But, we might also find that there's still a need for computers, as way of learning about the world, for games, for art, and for music, but once we identified where computation actually has a strengthening effect on the ecosystem, in most cases it will be used to do benign tasks." —Devine Lu Linvega (AKA Neauoire)

Some 'permacomputing-appropriate' solutions are direct and practical, such as passively illuminated screens, which refract sunlight to remain legible, dramatically reducing energy consumption (and improving optical health). In other ways, it presents broad prescriptions for how we conceive of and relate to technology itself, especially in consideration of a world that will likely become more energy scarce and resource constrained.

One framework described by programmer and artist Devine Du Linvega is that of ‘frugality,’ ‘salvage’, and ‘collapsability’. In short, that means we make the best use of existing resources (frugality); we re-use, repair, and reclaim materials to reduce waste and the need for new machinery (salvage); and use computers in ways that attenuate their utility and resource demands in situations of scarcity (collapsibility). When there is less energy available, for example, computers can simply run slower, or not at all, and that has to be ok.

These are design principles, and also a philosophy of relating what we use to the limits imposed by our circumstances on planet Earth. In light of the seemingly likely collapse of our highly energy intensive, supply chain-dependent status quo, the idea takes on a particular meaning, focusing on extending and maintaining resources, and eventually, considering how to meet collective and individual needs without tools that emerge from and rely upon unsustainable industrial systems.

In the event that large-scale electronics manufacturing were to suddenly halt, or a region were to be cut off from the global supply chain, computing devices would become precious resources whose functionality would need to be carefully maintained. Furthermore, without reliable power generation and distribution, long distance communication over the Internet as we know it today would likely not exist, even if the hardware itself could be maintained. Lack of connectivity would render all modern network-based services and software maintenance infrastructure defunct. Conservation of existing distributed hardware and software resources would need to continue until we either learned as a society to recreate their functional equivalents in a more sustainable way, or learned to do without them. — Esther Jang and co-authors

Software, too, is becoming bloated as hardware capacities increase. Functions that could be accomplished by simpler programs are now undertaken by increasingly complex and 'black boxed' software that eludes the understanding of more and more people — Ai is here, again, the example par excellence these days. None of this is sustainable, and runs in parallel with the broader death drive of extractive capitalism that is, if we're honest with ourselves, doomed to end one way or another, and probably sooner than most of us are prepared for.

The permacomputing perspective recognizes the immense social, creative, and productive value in the role of computers, but seeks ways to align them with ecological principles. There is no single definition of the concept, but it largely accords with principles of degrowth, and asks how computation can be ushered through the phases of collapse as a world order built on extractive industries putters to its inevitable end.

In permaculture, energy is not produced so much as captured and redirected; the latent processes of nature do the work, while humans guide it to desired ends in a more or less passive way. In doing so, those complex ecological and environmental processes are given wide enough berth to continue undisrupted, and something approaching a balance is achieved. These systems are effective and resilient not by virtue of any single set of engineered elements, but because of a range of components that interact in myriad ways. Is there any reason our 'technology ecosystems' cannot be similarly organized?

"In nature, everything is interdependent, and these interdependencies tend to strengthen the whole. In technology, however, large dependency networks and "diversity of options" often make the system more fragile. Civilization should therefore try to find ways of making technological dependencies work more like those in nature, as well as ways of embracing technological diversity in fruitful ways." —Ville-Matias Heikkilä (AKA Viznut)

At this point in history, it is difficult to envision a future without computers. That's a bit funny, given how recent these devices are, but we may also be confusing process with the means of achieving it. Computing in one form or another has been performed by human brains for millennia. One can perform a program on a piece of paper, albeit very slowly; even a bacterium can 'run Doom'.

None of the above should be incompatible with continued innovations that make computers better and more ecological. Technology will inevitably continue to evolve, and hopefully in a direction that is far less damaging to the environment. Computers are physical, after all, and we can build them out of damn near anything. Quantum computing for instance seems poised to introduce entirely new landscape of possibility for the form, application, and efficiency of computers.

"Right now our computers are so against the grain of nature," Linvega told me in a recent interview. "There's probably going to be computers in the future that are more akin to the laws of nature. It might be surprising to hear that I’m into permacomputing, but I also think quantum computers are the future, and I would love to have one and to be able to use them because at one point, I'm pretty sure running a program on a quantum computer is going to be way more energy efficient than what we use today."

Computers are all made of stuff we pull out of the ground and manipulate with energy. If the advancement of technology is a kind of civilizational software, it runs on the finite hardware of Earth itself. Bringing those aspects into sustainable alignment will prove necessary as resources grow more constrained, as the consequences of extraction and pollution and tenuous supply chains continue to mount. In thinking about that future, and where computers are or are not necessary—despite their ubiquity today—permacomputing offers some interesting and challenging provocations that are well worth considering.