At the start of 2020, Fast Familiar, Abandon Normal Devices and Arts Catalyst began The Networked Condition: Environmental Impacts of Digital Cultural Production, a collaborative research project exploring digital arts through an environmental lens. The project is part of The Accelerator Programme (led by Julie’s Bicycle and Arts Council England), which helps artists to advance their sustainable practice and share insights with their peers and the wider sector.
As part of The Networked Condition: Environmental Impacts of Digital Cultural Production, we’ve been speaking to artists and arts organisations whose work encompasses digital tools and environmental sustainability or activism, hoping to gain insight into a broad range of ideas and approaches. We hope that our case studies can be a source of shared knowledge, inspiration, or reflection for the sector too. You can find an interview with Invisible Flock here and one with Kyle McDonald here.
Here we chat to Memo Akten; artist, experimental filmmaker, musician and computer scientist from Istanbul, Turkey. He works with emerging technologies and computation as a medium to create images, sounds, films, large-scale responsive installations and performances. Fascinated by trying to understand the nature of nature and the human condition, he draws from fields such as biological and artificial intelligence, computational creativity, perception, consciousness, neuroscience, fundamental physics, ritual, and religion. He has a PhD in Artificial Intelligence/ Deep Learning and expressive human-machine interaction from Goldsmiths University of London and is Assistant Professor of Computational Arts at University of California, San Diego.
This is an extended version of the interview which can be found on Abandon Normal Devices' site.
Can you tell us a bit about your practice and what interests you at the moment?
I’m a computational artist, so I primarily work with computation. I write software – that’s the medium I work in. The biggest thing that I’m interested in is generally trying to understand stuff, and that ranges from trying to understand things at a very fundamental level, fundamental physics, so how nature works, to at the other end of the spectrum, on a societal level, how humans work as individuals and how society functions and all of the things in between.
Tell us about your research on the ecological impact of NFTs and machine learning art.
I work a lot with emerging technologies, and I'm interested in their impact on us as individuals and as a society. I’ve been very interested in AI (Artificial Intelligence) for a long time because AI is a means to enable computers to understand the environment that they're in.
I've also been interested in the blockchain since its inception - there's a lot of artworks I admire created using the blockchain, such as Primavera De Filippi's plantoid (a kind of autonomous virtual creature) and terra0, an autonomous forest.
I've been watching the lives of NFTs too, and did a lot of research that exploded at the start of 2021. I was concerned that people weren’t aware of how these Proof of Work blockchains worked - obviously a lot of people are, but I think many of the artists getting into the NFT space weren't aware of the ecological impacts of the “mining” Industry. It's not just in terms of emissions, it's also e-waste. I think the Proof of Work (PoW) industry is an abomination, the absolute worst of humanity and I don't think artists getting into this space had any idea. They thought uploading an NFT was maybe like tweeting or putting a video on YouTube or Tumblr or Instagram and it's not.
So, this research started purely out of my own curiosity because I was also thinking about getting into this space - it's very alluring for many reasons. My practice relies on me traveling, I work with festivals giving talks, which means I need to fly a lot. I wanted to cut down on my flying, so I looked into NFTs as an alternative and couldn't find any figures about carbon footprint. No one had done this research and so this led me down many rabbit holes and I finally wrote that report in December 2020, which exploded.
A lot of people were not very happy. There was a lot of push back and a lot of ‘whataboutism’. And for quite a few months the conversation that I wanted did not happen. People were either violently against me, or violently with me, but the discussion wasn't happening. It was just, ‘Well, money is evil, so how can you use money?’. Now, a few months later, we're able to have the conversations that we need to have about it.
Why do you think people found it so hard to accept?
One of the most common pushbacks is ‘yeah, but do you fly?’ or ‘yeah, but do you drive a car?’ or ‘but do you use money?’. And immediately people read the report as criticism, as me saying “I'm better than you”. It's seen as a personal attack and it’s very difficult to divorce that point of view from the overall conversation. We have a similar situation with regards to eating meat and animal products, for example. I am not a vegetarian, but I do think it's wrong to eat meat, given that the whole animal industry is absolutely abhorrent. I believe our descendants will look back at our civilization and the way we treat animals with the same level of disgust with which we look at our ancestors who had slaves or went to mob lynching or to hangings for fun with their kids.
I always think, “OK, you criticize me for using the US dollar, but that doesn't make the Proof of Work industry any less exploitative”. So, I think this is the root of it: how to make it clear that this isn't about ‘I am better than you’, but ‘let's talk about this’. The tool that I built is essentially for calculating the carbon footprint of something, but it became weaponised in a way others didn’t.
One thing I found interesting was the use of the tool that I built. In retrospect it's very easy to look at it and say “it's a tool for shaming” because that is what it was used for, and I had to shut it down because it got out of control. But as I said at the time, we have carbon footprint calculators for flying and yet we don't have a culture where if someone tweets ‘I just landed in Montreal for a music festival’ people reply to them with a screenshot saying ‘you just did 1 ton of CO2.’ We don’t do that, but it happened with NFTs. Just because there's this tool for calculating the carbon footprint of something, why did it become weaponised, when other carbon use didn't?
What I wasn't aware of when I launched the tool was that there was already so much pressure, such anti-NFT sentiments brewing, that secretly many people already hated NFTs for various reasons, for the hyper capitalist aspects of it. Then when I released that tool, the anti-NFT movement exploded, and the tool became the weapon for the anti-NFT movement… that’s my hypothesis.
And what about the wider societal impact regarding climate change? How has this research impacted your own actions, separately from the awareness-raising you were doing in the tool?
One of the big push backs that I keep getting is that ‘individual actions don't matter’; the idea that what you or I do has no effect, which is one of the most fatalistic lazy points of view one could have. The tragic thing is, there is some truth in that statement, but the truth in that statement is from an individual numerical point of view - and only if we ignore the impact of our actions on others. What I do, what you do, is obviously a drop in the ocean. But what matters is with regards to the pressures that we put on businesses and in a way on each other, because collectively it's the mindset that makes the difference.
So, whether I make one less flight or not, it doesn't actually have an effect directly, but the effect that it has is on my environment and those are the people saying that ‘I care about this’ and ultimately it can propagate out to businesses whose agenda is to maximise profits. The CEO of a company has a legal obligation to their shareholders to maximize profits. So, with all this in mind, we have to be aware of these market forces and that was really what was behind all of this.
NFT businesses like SuperRare, Nifty Gateway, etc. didn’t just accidentally pop up; they were all funded around 2017, startups with Venture Capital (VC) funding. So, this whole NFT market is actually built on crypto investment agencies wanting to bring more people into crypto currencies to pump the value of their assets. It's not that if you mint one less NFT, you're going to save a ton of CO2, but it's the signal that we send to businesses and each other in terms of what we value and how businesses should act.
Has it changed you as an artist in terms of the work you create, or rather is this research something that sits alongside your practice?
I've definitely learned a lot. I don't know if it's influenced my work yet, but I think it definitely will and it’s not the research itself that I've learned the most from, but the reception to it. In a way it has been very sad seeing so many people dismiss my criticism based on essentially, ‘but I'm making money’, ‘but it allows me to make a living’ and that is never a moral justification. Observing that was very eye opening and it will affect the work that I make. It just hasn't yet because I need time to let it sink in.
If an artist were to ask your advice in creating digital art or NFTs in the most sustainable way possible, what useful tips would you give them?
This is a very tricky discussion. It's a super speculative hyper capitalism turned into an algorithm. I understand many people would still object to NFTs, even if the ecological issue wasn't there, because right now there aren’t many alternatives, for example, the primary platform in use is Ethereum, a stupendously energy-consuming blockchain.
Some new, alternative blockchains are emerging with much lower energy consumption, but many blockchain purists don't see those as “safe” or as truly decentralized because of the way that they work. It’s true that these Proof of Stake blockchains are also very problematic, not for ecological reasons, but for other egalitarian reasons - they effectively help rich folks get richer. However many points you have, you get given more coins proportional to that, so it's a computisation of real-world market dynamics. If you object to that, then there is no way of getting into NFTs; crypto currencies in general are inherently hyper capitalistic. Unfortunately though, if an artist sells on an Ethereum blockchain they make at least ten times more (if not hundreds of times more) because the high-end collectors seem to want Ethereum tokens more than they want the art.
There are, however, many new and emerging platforms that don't use Ethereum. I've recently started using Hic et Nunc. It’s on a Tezos blockchain, open source and community run, so there's no big company behind it. This also means that it's buggy (up and down, crashing). I wrote a guide listing a number of eco-friendly NFTs in April 2021 which is a great resource (also free to explore on discord).
Do you have advice on approaches to machine learning arts and Generative Adversarial Networks (GANs)?
Machine Learning models have undergone a very interesting transformation in the last few years. Just a few years ago - 2016, 2017 - you could train GANs on a laptop in a day or even a few hours. The GANs that I used the most (including for my most well-known work Learning to See) would take 2 hours to train on a gaming laptop. And this is 2017. It’s like watching a film, two hours, nothing in the grand scheme of things - nothing even, in the grand scheme of your own emissions. Now, GANs like StyleGAN can take four months to train, if you use a high-end GPU (Graphics Processing Unit). I don't really train those kinds of models anymore - not just for the energy, but also because it's expensive.
My advice would be to think about what it is that you want to do. I should also add that we're getting into a new realm of Transformers, which you can't train with any kind of consumer hardware. Transformers (or really big models) use hundreds of TPUs (Tensor Processing Units) which means we're reaching an interesting stage in Machine Learning where maybe five years ago, an independent researcher could train the same models that the high-end researchers were training, which is no longer viable. However, we now have what some people call foundation models. Some people find that a problematic term, but the idea is that there are these huge models which are trained likeGPT-3 and then people like you or me can fine tune them, within minutes. This is interesting because the effort that goes into training a foundation model is shared between the millions of people who end up using that model. So, you can take a model that's been trained and then just fine tune it. And this really is hopefully the way forward, not just for ecological reasons but also because it makes more sense, especially for artists who don't want to tie up their computer for four months only to realize that the model didn't come out well.
What first made you think about researching the ecological effects of your work? You mentioned that a lot of your interests were rooted in nature to begin with, I wondered if that was the reason that you made that connection, or did it come from a different place?
When I say I'm interested in nature, I don't necessarily mean the countryside, but I mean the laws of physics, like ‘the nature of nature’ so to speak (I do also love the countryside). I researched the NFT stuff because I knew that the Proof of Work algorithm that Bitcoin uses is incredibly… I'm going to say wasteful, and many people disagree with that. They say it's not wasteful because it's creating value, basically it's converting electricity into value and there's all these justifications. But I had no way of connecting the overall picture of how the Proof of Work industry works to an NFT. It was very interesting research and it consumed six months of my life in total. And it's not over, because the idea of a carbon footprint for a single NFT is very complicated. It is complicated to measure the carbon footprint of the entire mining industry. But at least with that, there is an objective truth out there - I mean, we won't be able to measure it, but there is a truth to it. A bit like if I were to burn a ton of coal right now, then that is going to emit a certain amount of CO2. There is a certain amount of emissions related to a particular flight, but if you're sitting on that flight, what is your footprint? What is your contribution to that? This is open to debate. There are rules of thumb as to how this is calculated, how that responsibility is divided up. But it's obviously not necessarily objectively true. Do you do it by your weight? Do you do it by how much area you take up on the plane? Do you do it by how many people on the plane and you divide by that, or do you do it by how much you paid? Because obviously the airline is a profit-making company, so the more you pay, the more incentive you're giving for them to fly. So, there are all these questions and we have similar and actually, even more complicated discussions around the carbon footprint of NFTs, because, again, one of the common pushbacks is ‘mining happens whether your NFTs are there or not.’ Which is a bit like saying ‘planes fly, whether you sit on them or not.’ It's very complicated to assign a footprint to an individual.
Do you think that people are quite happy to use carbon-intensive processes, partly because it's so inaccessible to calculate your carbon footprint?
Human psychology is really fascinating. People want to create NFTs because they're making a ton of money. We're not just talking small amounts, we're talking really big amounts. So the attack on my research is that “oh that figure is not accurate” and then, “oh that figure is not accurate, so I don't believe any of it, and it isn’t really a problem.” Whereas yes, it’s not accurate, but it’s still in the right ballpark. And that ballpark is that a single NFT has a footprint similar to that of an international flight. I forget the exact figures (it is in my article), but a Tweet is something like a few grams of CO2. Or even watching an hour of Netflix, is 10s of grams of CO2, whereas an NFT is tons, so it's still so many orders of magnitude different. The fact that the figures I quote are not accurate shouldn't affect this conversation. As a distraction, people prefer to argue over the precision, as opposed to the general magnitude. But that doesn't matter. As long as you can attack it, people say ‘Oh, it's not accurate anyway’, and then they're OK with the emissions. It's been very interesting to observe this.
One of the craziest defenses that people get behind is the idea that carbon footprint was invented by a BP advertising campaign, to put responsibility on the consumer, so we shouldn’t take it seriously. BP didn't invent the idea of a carbon footprint, they simply popularized and used the idea of a (personal) carbon footprint to put pressure on the consumer so that they themselves could evade responsibility. (As a side note, note that my tool is not calculating a *personal* carbon footprint, but the footprint of a *product*, a concept that goes back to at least the 1960s). As long as there's something to hide behind, people will hide behind it. As long as the door is left ajar for people to disagree with the real impact, they're OK with it. It's classic politics, really and it's been very sad to observe it.
Could you explain why mining has such higher orders of magnitude impact than watching Netflix? You mentioned e-waste earlier, why does e-waste result from it as well?
Proof of Work is one of the ways in which a blockchain is secured. A blockchain is effectively a decentralized database. Let's say you know your bank has a server that keeps track of everything, or Google has multiple servers that are all synchronized - but one entity owns these servers. In a blockchain - and this idea is actually very noble - no one entity owns the server, everybody runs servers and we all have a copy of the database. But then how do we prevent me from changing my copy of the database and saying, ‘hey, this is how the database should be’? This is a problem that needs to be solved. One method that Bitcoin introduced actually predates Bitcoin by decades, but Bitcoin combines it to the currency. They are able to write the database you need to effectively find a magic number and there's no way of knowing what that number is, other than generating a random number and then hoping that it matches. So, you have to generate a random number and put it through some algorithm to compute the hash and then see if it matches what the algorithm is expecting. So this is how one writes to the database. If you get it right, if you generate a random number and it matches, you get a reward, and that reward is basically their currency. Because it's become so profitable now, there are millions of miners around the world who are generating random numbers trying to find that magic number, the key, or the hash that's going to match the next block to be written. If you want to make a transaction, these transactions get routed to the blocks and for that block to be written to the database literally millions of computers around the world are just sitting there generating random numbers all day, trying to find the right number for the next block. In the case of Bitcoin, a block is written roughly every 10 minutes. For Ethereum it is every 10 to 15 seconds.
E-waste is created because it is so profitable. For one thing, there is specialized hardware that is being made at ASIC. Mining is very competitive because you have millions of miners around the world all trying to find the same number. If someone finds the number before I do, they get the reward and I get nothing, so it's in my interest to find the number before they do; to invest in faster hardware to find it, and new hardware is constantly coming out. So that means the equipment I bought last year is no longer profitable. There's a very interesting dynamic when the price goes really high - too, as it might become profitable to turn on some of these older mines. With Ethereum, the algorithm doesn't work with ASIC, so GPUs are used for mining Ethereum, and in fact there's been GPU shortages. Prices have been going up, there's been a short supply of GPUs because they're being used by Ethereum miners and particularly the ASIC, they can't be used for anything else. They use an ASIC for a year, and then a year later it becomes obsolete because a new ASIC comes out and then they just get dumped, likely because they're not general purpose computers. The only thing they do is mine Bitcoin.
So why, then, are the smaller community blockchains cleaner?
They use a completely different algorithm. With Proof of Work, the basic problem that needs to be solved is, here's a block of transactions that we want to write to the database, and because everybody has a copy of the database, when this block is sent to everybody, everybody needs to be sure that the block is a genuine block. So, the way they solve this is through Proof of Work, with these miners that are trying to find this random number. What Proof of Stake does, is to use Stakeholders, who are people with lots of coins of that particular cryptocurrency, who vouch for the legitimacy of that block, the idea being that they wouldn't do something fraudulent because if they did, the value of their coins would go down and they would lose a lot. The basic premise in Proof of Work and Proof of Stake is that to attack the network and produce fraudulent blocks would not be profitable for those who are trying to do it. In Proof of Work blockchains this is accomplished through the idea that to be able to produce a fraudulent block, you would have to invest in so much hardware and so much electricity that it would not be worth the benefit from producing a fraudulent block. And in the case of Proof of Stake, it's based on the premise that you hold so much stake in the network value as is, that if you produce a fraudulent block, you would lose the value that you hold.
Why do you feel people are more open to the conversation around the environmental impact of NFTs now? Are you feeling more positive or more hopeful about us being able to talk about this more openly?
I think it comes back to the personal attack aspect. A lot of people said that I was shaming them right at the very beginning, and I find that very interesting because bullying someone is one thing, but shaming? What that says to me is that people feel ashamed. I think there's a lot of cognitive dissonance. When I wrote the article, it was followed by this huge online backlash as well as in people who were anti-NFT and started attacking and lots of things like that. I think now some of that has subsided, and so now we’re not taking things so personally, we can and talk about how to move forward. I'd like to believe that the popularity of some of these non-Ethereum blockchains is in part due to this research and this conversation, because otherwise why would people take a massive pay cut to switch to these platforms? They would just stay on Ethereum. But I still think we're very far from actually solving the issue.
I think the most important thing to get out there is to counter this myth that individual actions don't matter. It's not a binary, it’s not ‘they matter /, they don't matter.’ There’s nuance to it. Individual actions don't matter purely from a numerical point of view, but values and actions are contagious, and so, individual actions do matter in the grand scheme of things when you think on a societal level. I don't litter and say “oh it's just one bottle” -, that’s a stupid way to think. It's a tricky, tricky subject because it's such a big, big problem. Ultimately, Climate Change is the problem that we face and that’s where all the roads lead.