-- Leo's gemini proxy

-- Connecting to zaibatsu.circumlunar.space:1965...

-- Connected

-- Sending request

-- Meta line: 20 text/gemini

Discussions toward radically sustainable computing


Approaching three years ago now, toward the end of 2017, my still young phlog went through a period of still unrepeated thematic coherence when I spilled a whole lot of bytes on the topic of radically simplifying one's life, and selectively purging it of a lot of modern technology, in service of multiple goals but chiefly reducing environmental impact and increasing available leisure time. All of this was written shortly after my move to Finland, at a time which I perceived the be the possible start of a new chapter in my life. I thought hard about this stuff and had plans to make concrete changes to slowly but surely push my life in a particular direction. Representative posts are:


Lithium blues (2017-10-10)

More battery thoughts (2017-10-11)

Asceticism, or something like it (2017-10-19)

Technoskepticism, or something like it (2017-10-22)

Radical frugality, or something like it (2017-10-24)

Frugal entertainment (2017-10-26)

Hobbyist electronics without hypocrisy (2017-10-31)


Shock, horror, I never wrote about any of this stuff again with anything like the same amount of energy, frequency or consistent focus, and while it's not true that I completely stopped making the kinds of efforts I wanted to make, it's fair to say that I fell pretty short. In particular, instead of carefully "reforming" my electronics hobby, I more or less entirely dropped it, partially as a consequence of increased cost and difficulty of acquiring supplies in my new location, but more so, I think, as a result of getting distracted by an unexpected but intense love of outdoor cycling and accompanying bike tinkering, and also by the whole Circumlunar Space thing taking off.


Fast forward to the present and I have recently started having increasingly frequent, engaging and rewarding discussions with people in the Fediverse on the "sustainable computing" chestnut, which has brought me back around to this old train of thought. I am getting somewhat fired up on the subject again and I think I want to resume seriously thinking and writing and maybe even gosh-darned *doing* something about it on a semi-regular basis. Since I first wrote a lot on this matter I suspect that:


I've gained some new readers, especially from Geminispace, who never read what I previously wrote

Some of my old readers have forgotten a lot of what I previously wrote

I myself forgotten a lot of what I previously wrote

My ideas have changed a bit in some respects


Thus, I am kicking things off with a characteristically verbose wall of text more or less "setting the stage", outlining my current thoughts and beliefs in this sphere and setting a rough direction for where I want to take the conversation, which I hope is something that people will find interesting and want to participate in themselves. Compared to many people interested/active in this sphere, I suspect a lot of my opinions and motivations will be perceived as extreme, if not quite radical, and I don't expect everybody to agree with everything. I am not dogmatically convinced of anything I'm about to write, I am open to hearing alternative points of view and I might even change my mind on some things, but what follows is where I'm starting at right now and where I think the discussion should go from there. Let's begin!


I am fairly concerned in my day to day life - certainly not as deeply as some people but more, I believe, than average - with issues of sustainability. This is despite the fact that I am actually pretty pessimistic about the whole prospect. I am not a dyed-in-the-wool "Doomer", and I'm not actively prepping for the imminent collapse of civilisation. I would not even go so far as to say that I am convinced a collapse *is* coming. But I consider it a very real possibility, and I think anybody who dismisses it out of hand has certainly misplaced their confidence. Even if collapse may still be in-principle avoidable at this point in time, I am not remotely optimistic about the prospect of humanity successfully executing whatever evasive manoeuvres may be required.


Never the less, I do try fairly hard to reduce my environmental footprint in whatever ways I can, and this is an ongoing journey that I'm pretty committed to. I'm committed to this even though I'm aware my individual efforts are a proverbial drop in the bucket, because I simply don't seem to have it in me to do otherwise. I am not one to fiddle while Rome burns. Even if it won't make a difference, even we are all doomed, I find it very hard not to "do the right thing" once I have formed some idea of what I think that is.


I also make the effort to actually exercise good environmental practices myself when I can because I'm on the whole very big on direct, bottom-up individual action being the best way to fix things rather than top-down salvation from governments or charities or companies or anything else like that. There is a prevalent narrative online these days that "ordinary people" bear no moral responsibility for climate change or other impending environmental catastrophes, that "our leaders" have let us down with inaction and that "the corporations" are really to blame. I reject this narrative entirely: it is a cop-out. This is not to say that politicians and corporations are blameless, they surely aren't. But corporations aren't destroying the planet for fun out of the evilness of their hearts, they are doing it incidentally as a side-effect of supplying goods and services that, for the most part, "ordinary people" willingly consume and which enable a lifestyle they would actively resist giving up, and politicians don't have a magic wand they could wave to make that lifestyle sustainable. The problem is just not that easy. Real, dramatic change in the way each and every one of us lives is a non-negotiable requirement for actual long-term sustainability.


Not only is individual action thus a moral imperative, I also believe it often has lower barriers and is more secure against later reversal than political action. Not only do no new laws need to be passed, overcoming overwhelming opposition from entrenched commercial interests, in order for me or you to reduce or eliminate our meat consumption, and hence substantially reduce our carbon footprint, but in most parts of the world it's unthinkable that a law *could* be passed, even by the most corrupt and delusional government, that could actually stop us from doing it. No government is going to *force* you to eat meat. This is just one positive change of many which can be made immediately and permanently as soon as you have the conviction to make it, with no dependence on outside authority, which is inevitably susceptible to abuse, corruption, its own collapse, etc. How is that not the best way to address any problem? When so few of us - myself, let me be clear, included! - have managed to adopt these changes, after the situation has been well publicised for decades, how can we blame higher ups for letting us down? How can we still genuinely expect that the higher ups will actually suddenly start doing something in the future? We can, and must, start solving the problem ourselves using the unassailable powers we each have, by aggressively opting out of as many of the drivers of destruction as we can, as soon as we can.


I think it is delusional, or at the very least unduly optimistic, to think that anything even slightly resembling a modern Western lifestyle can be made sustainable for the entire current, or near-future projected, global population simply by switching out all the familiar parts with new, greener versions - replacing all the petrol cars with electric cars, replacing the coal power plants with solar or wind ones, etc, etc. It's not that those things don't help at all, and I'm not saying we should actively resist them, but a lot of their impact will be negated by the Jevons paradox, and besides that this is simply far too minor a change, a failure to recognise the full scope of the problem. To some extent, climate change and the issue of carbon-neutral generation of electricity have "stolen the stage" of sustainability issues in the public consciousness, and are stopping us from confronting the broader implications of finite resources. Peak oil is only the beginning! Every aspect of our modern life has non-renewable resources underlying it that will *eventually* run out if they are consumed without limit. Peak lithium will be a thing one day. Even peak copper and peak steel will be things. We can't *make* any of this stuff, we can only refine and use stuff we dig out of the ground which which has been there since the planet was formed. Some of it can be recycled and reused (with an energy cost, of course), but the total amount in circulation is fundamentally capped and no plausible technological innovation in the near future can possibly change this. Even if you can convince yourself that all these peaks are millennia into the future (which is still not the same thing as never, of course), the issue remains that acquiring these resources and rendering them useful is unavoidably destructive. Even if all your mining equipment and refineries are happily solar powered, we are still talking about cutting down forests and digging holes in mountains to get raw materials, and refining them via chemical reactions that produce nasty byproducts. When you admit that just about everything is inherently and inescapably finite, and that even the most careful use of those finite resources involves non-trivial environmental damage, simple intellectual honesty immediately leads to the conclusion that genuine, long-term sustainability requires using as little of just about everything as possible, and being extremely careful about what you do use. Never ending improvements in convenience and material comfort are strictly incompatible with this. Most mainstream ideas of sustainability take little more than few small steps in this direction.


Jevons paradox, for the unaware


So, my vision of sustainable society looks an awful lot closer to pre-industrial society than it does to modern society. I don't see any need to go quite to the extremes of anarcho-primitivism. I think there are technologies which are sustainable - or sustainable *enough* - relative to the amount good they do - that they are worth preserving. But a lot of it will simply have to go, and I think we really ought to begin moving in this direction hard and fast, undergoing rapid "degrowth". Large governments and large corporations are not ever going to champion this idea themselves, but that's okay because they don't need to: we can do it, or at least most of it, ourselves. I certainly don't *enjoy* every part of this prospect, and I don't think I have a romanticised vision of "life in the good old days". Some of it will be unpleasant, no question, but given the choice between "unpleasant" and "systemic ecosystem collapse" I don't think the choice is difficult. We simply have to roll a lot of "progress" back. But it's very important to avoid thinking in terms of a literal return to exactly the way life was at any point in time during our recent past. Returning to any historical level of resource consumption from the present will be extremely different from the first time around. I wrote a little about this previously, using the language of a "mixed timeline". This time, the amusing tagline of "pre-modern life with modern characteristics" springs to mind as an alternative. There are two sides to this: intellectual and material.


Intellectually, there is no reason that actively moving away from the energy and resource consumption of global, industrialised capitalism means we need to give up any of the scientific knowledge we gained under it, or reverse any of the shifts in social attitudes which happened at the same time as (perhaps facilitated by, perhaps not) its reign. One of the reasons life sucked "in the good old days" was not having the germ theory of disease. Well, we have that now! And we don't necessarily need to lose it if we start living a more materially "primitive" lifestyle. Nor would we need to forget everything we have learned about health and medicine in general, or agriculture, or nutrition. A concerted effort to "live the old ways, but smarter" could produce very different results to the first time around. Nor would we need to give up the commitment to gender equality, anti-racism, or any other such social progress. We could "live the old ways, but fairer".


Materially, though, is where things really get interesting, at least for me. I believe a sustainable future necessarily involves far, far fewer and far, far smaller mines, refineries and factories than the present day, but even if we started shutting and/or scaling down those things today, that doesn't mean that all the things we have manufactured to date would suddenly and spontaneously disappear. You know how Cuba is full of classic 50s and 60s American cars imported from before the Cold War embargoes, and maintained ever since by increasingly creative and resourceful mechanics who keep them running with a hodge-podge of available parts? That's the first century or so of a genuinely sustainable future: using, repairing and adapting pre-existing artifacts from "the before times" which we no longer allow ourselves to manufacture new instances of, for as long as we can, getting the most and least destructive use out of them we can, while we gradually wean ourselves off our old ways.


This realisation opens up *so many* fascinating questions: of all the high-tech stuff we are mass producing today, which can only be produced by wreaking terrible damage on the ecosphere, which things would remain useful if all the trappings of modern industrial society disappeared, and which would immediately become useless? Of the stuff that would become useless, can we think of clever ways to repurpose it, or reuse its parts? Of the stuff that remains useful, what will fail fastest, and what are the prospects for repairing it using scavenged parts and tools? What would be the most useful spare parts to stockpile now while we can? Can we make modifications to stave off the most common failure modes? Of the stuff that we think we might actually have a hope of keeping running for a long time, what are its most important applications in a genuinely sustainable world?


If you want to be optimistic and assume that we can afford to put a slow and gradual end to rampant industrialism, and that we as a species can get our shit together enough to actually *do* that in a controlled way, then the answers to these questions would be invaluable for guiding our decisions on how to make use of the final few decades of destructive manufacturing. If we're not going to stop churning stuff out immediately, we need to at least immediately start optimising the designs of everything we *do* make for repairability and adaptability, and to start manufacturing and stockpiling replacement parts. So, both pessimists and optimists alike have motivations to think about this stuff, which I hope means my ponderings on these matters will be of broad interest.


Finally, we arrive at computing, or perhaps consumer electronics a little more broadly. This is nothing more than a special case of "high-tech stuff" as discussed above and everything said there applies here. I fully grant that it might seem like a strange special case to dedicate attention to - when all is said and done, if we're talking about stripping modern industry back to the bare minimum necessary to provide for happy, healthy livelihoods, computers just don't really seem to make the cut, even if we're being generous. Computers were not essential to developing antibiotics or vaccination or blood transfusion or a whole bunch of other technological breakthroughs that drastically improved people's lives. Within living memory, millions of people lead daily lives not touched by computing in any way, and while those lives may not have been perfect, it would be hard to argue that a lack of computation was the main contributing factor to that imperfection. If we were talking about designing a long-term sustainable lifestyle to be implemented from scratch, it seems to go without saying that computers simply wouldn't figure into it.


But, of course, that's not what we're talking about. We are talking about the possibility of maybe just barely being able to rapidly transition to a long-term sustainable lifestyle starting from the world we're in right now, and that world is absolutely lousy with computers. Even if we shut down all the semiconductor fabs tomorrow, those computers wouldn't go anywhere. The overwhelming majority of the environmental damage each of those computers represents is already done, by virtue of their having been made. The ongoing energy requirements to keep them running are indeed very, very small, especially the modern single board ARM machines, and absolutely within the reach of what a radically decentralised, off-grid, sustainable eco-village could generate itself using scavenged technology. Having some kind of minimal computing infrastructure in place during the transition from the present to a sustainable future would surely be useful. I can't shake the feeling they could play a very important role in that whole "not forgetting all the very important technical/practical know-how we accumulated before we got ourselves under control" thing.


Above and beyond the issue of non-stop destructive manufacturing of electronic devices with ever shorter useful lifespans is the matter of the environmental footprint of the modern internet infrastructure. It's very hard to imagine that high speed, low-latency, 24/7 connectivity between just about any two points on the Earth's surface can be made sustainable. The question of how we move away from that vision of the internet is no less fascinating than how we get the most and best use out of existing hardware once we know nothing newer will ever come along. Within a community of sufficient population density, existing hardware and software is already perfectly capable of implementing WiFi and/or Bluetooth mesh networks with zero dependency on any infrastructure external to the computer themselves - no cables, no cell towers or antenna masts, no satellites, no datacentres, nothing - so networking need not disappear entirely. But communication *between* such communities, across expanses of insufficient population density for meshes, is another question entirely. I believe it will be possible, but also that it will of necessity be very slow, very unreliable, and very intermittently available, by the standards of the modern internet (which, of course, can still mean mindblowingly fast and reliable by historical standards, like people carrying written messages around on horseback or in boats). When low latency is required, ideas from amateur radio (particularly packet radio and ultra low power "QRP" operation) will probably be highly influential. When latency is not important we'll resort to "sneakernet" methods. The old joke says "never underestimate the bandwidth of a station wagon full of tapes hurtling down the highway". There will be no functioning station wagons in a sustainable future, but thanks to dramatic advances in storage technology, a backpack full of 1TB or larger USB hard drives is small and light enough to be carried long distances by a cyclist, providing bandwidth that's certainly not to be sniffed at, especially when its use is limited to *important* information, stored in the most efficient way possible. How, and for what purposes, can we best make use of an "internet" with these properties?


Even if you think I'm, frankly, batshit insane with all this ecological collapse and radical deindustrialisation talk (I don't deny it's possible!), I hope you'll still agree that we - where "we" refers to society in general, but in particular to us, "the computer people", who have made these wonderful things central parts of our livelihoods, education, recreation and socialisation - are kind of overdue for frank and open discussion about the environmental impact of computing and the internet, especially as the market shifts ever further away from a model of people owning a small number of comparatively large, repairable, upgradable devices used for many years toward one of people amassing veritable fleets of smaller, mobile or even wearable devices not designed with longevity or generality in mind. And, of course, especially as the Internet of Things rears its ugly, ridiculous head and promises a future life where everybody's house has dozens of even hundreds of plainly unnecessary tiny computers in it, each one containing rare Earth metals, many of them containing lithium batteries. Even if you don't think we need to go to the same extremes I'll be talking about in future posts, you'd have to be deeply in denial not to agree that *something* needs to change. Once again, I think both the optimistic and pessimistic versions of this discussion will have overlap that I hope even people who *do* think I'm nuts still think I'm at least worth listening to some of the time.


I think that, at last, is pretty much where I currently stand on these matters and the direction in which I am thinking of heading. I absolutely didn't arrive at this position on the strength of my own thought. I have taken information and inspiration from many sources and people. A woefully incomplete list of these follows.


The issue of embodied energy in computing hardware, and the energy consumption of the internet infrastructure, both came to my attention first through articles in Low-Tech Magazine:


The Monster Footprint of Digital Technology

Why We Need a Speed Limit for the Internet


xj9 of sunshinegardens.org wrote the post below (originally on the web but which I am so happy to say is now also on Gemini) about the monstrous energy consumption of internet infrastructure, and the notion of lower power, lower tech decentralised alternatives based on meshnets and sneakernets. It is written less from a perspective of environmentalism and more one of driving internet cost down to the point that it's effectively free in order to make it more accessible, but the same ideas serve both goals to a large extent. In general, xj9 has done a lot to inspire and motivate me, even though I think she's more optimistic and less primitivist than I am on the whole, mostly because she endorses the "Walkaway" philosophy (inspired by Cory Doctorow's novel of the same name - which I enjoyed but question the real world validity of) of individual actors opting out of the worst parts of modern society, which I think is a hugely undervalued technique, and also because she is big on salvaged electronics.


Universal Basic Internet


Devine Lu Linvega (@neauoire@merveilles.town) and partner Rekka (@rek@merveilles.town) live and work (as developers, artists and authors) onboard a sail boat, which has given them great real world experience and insight on the realities of low power computing and intermittent network connectivity. Over years they have migrated away from workflows based on power-hungry, update-dependent Macbooks to using multiple single-purpose Raspberry Pis wherever possible, and writing many of their own minimalist software tools for their work. However, in a wonderful article on "Longtermism", Devine acknowledges that "pushing for the purchase and production" of more "small low-power open-source single-purpose boards" is not without problems, and wonders if "creating software targeting old hardware might be what I'm looking for", which again aligns a lot with the salvaged electronics approach which I think has to be the fundamental basis of genuine sustainability.


The "Longtermism" article

The journey toward low-power and self-reliant off-grid computing

A fleet of Raspberry Pis, optimised for fixed purposes


Thanks for reading this far, if anybody did! I'm looking forward to having a lot of interesting and challenging discussions on these issues with various people in the Fedi-, Gemini- and Gopherverses in the future. Shufei (whom I've had some Fediverse and XMPP chats with about this) has already made a gemlog post on the matter, and I know there are other folks planning to chime in, too. I can't wait!


Shufei's 2020-07-05 post "(Compy (Solderpunk (Solarpunk)))"


Responses to this post


"The future internet", by Tom of Foobaz



-- Response ended

-- Page fetched on Thu Mar 28 22:24:56 2024