What happens when you introduce physical decay into an otherwise pristine digital environment?
Erode: breaking down NFTs
NFTs got a lot of attention recently. Was it a fad born of lockdown boredom? Or is it the future of digital art? Instead of judging from the sidelines, we decided to jump right in. Presenting ‘Erode’, a project that uses NFTs to critique the very idea of… NFTs.
This is not the essay in which we explain in precise detail what NFTs (Non-Fungible Tokens) are; there are quite a number of good explanations out there. Nor can we tackle every decent critique of the phenomenon, particularly its serious carbon footprint problem. You are probably aware that some people have made a lot of (real-life, fiat-currency) money from NFTs. Sir Tim Berners-Lee (he of WWW-inventing fame) has even sold — some would say quite ironically — “the source for the world wide web” as an NFT at Sotheby’s.
Depending on who you talk to, NFTs (and the related technologies such as blockchain, cryptocurrencies and decentralised applications) are either “Crypto’s latest pump and dump scheme” or the future of the Internet, Web 3.0.
But while we enjoy some healthy intellectual debates, our instinct as Random Studio is always to move beyond theory and into praxis. Instead of just talking or writing about NFTs, we decided to play. (Safely, in a sandboxed project running on an Ethereum testnet to avoid burning insane amounts of money — or fossil fuels.)
Our project, Erode, was a chance to learn about NFTs, to familiarise ourselves with the tooling, the challenges and the possibilities. As with all emerging and complex technologies, there are plenty of arcane concepts that need to be mastered and some tricky “unknown unknowns” which are only revealed through practical experience, so it’s great to have some of this knowledge on our team as a result.
But we never wanted to simply join the NFT hype train. We had to try something new.
On a practical level, we noticed that the dominant form of NFT artworks were either still images or occasionally short animated works (usually GIFs). Given our own interest in metaverses, digital twins, augmented reality and 3D modelling in general — wouldn’t it be interesting to explore 3D digital assets as NFTs?
There are plenty of practical challenges with trading 3D models, for example the proliferation of file formats, incompatible standards and inconsistencies in rendering. We settled on some pragmatic choices: glTF (intended to be an interoperable format for 3D content across industries) and rendering in the browser via Google’s model-viewer web component (which includes AR capabilities).
More importantly, we wanted to experiment with the medium of the NFT itself. This is the core of the erosion concept: we implemented an automated process whereby each time one of our 3D NFT “artworks” changed hands (i.e. transferred between “wallets”, which is fundamental to NFT trading), it would degrade. In this way we could introduce something analogous to a physical process into the otherwise pristine digital environment.
One of the curious implications of “minting” an NFT is that it imposes a kind of artificial scarcity on a piece of data that would otherwise be trivial to duplicate, transfer or modify. Erode would introduce a second layer: artificial “weathering” of a work that would ordinarily appear immune to time and damage from being handled too much.
Indeed, the mere existence of the NFT phenomenon calls into question the notion of “ownership” of a digital asset. What exactly have you bought? A claim on one piece of data (typically the “smart contract”) which in turn encodes a claim on another piece of data: the “asset” itself — typically residing on another, unrelated storage system. Erode intentionally reminds us of the buyer’s anxiety by degrading the supposedly “immutable” underlying asset in the very process of purchasing it.
We also needed to call into question whether keeping an artwork in pristine condition was necessarily an unalloyed benefit of digital art. The Night Watch can be justifiably restored by the Rijksmuseum because it is by definition the only one of its kind, and another layer of paint would presumably not change its claim to be the “original”. But would using AI to return it to its original size be a step too far, if the addition was permanent? And do we lose something of the appeal of a beautiful ancient ruin when we attempt to reverse the effects of time, as might be happening with the Acropolis?
Our erosion process was carefully tweaked to produce results that are aesthetically pleasing and interesting in their own right. The 3D model is put through a modified gltfpack “optimisation” process — normally intended to preserve essential details while reducing file size. At each degradation step, we keep the highly detailed HDRI environments intact and allow them to reflect off the increasingly abstracted/deformed model — an ironic treatment of the notion that digital art is assumed to exist in a void that has no relation to the external world.
Ultimately, we don’t think all “creative destruction” is automatically positive, either. By slowly destroying the NFTs in Erode, we also wanted to evoke the potential dark side of NFTs and “crypto” technologies in general. While trading our digital currency for digital items, and imagining it oh-so-ephemeral, are we not contributing to the destruction of a livable planet, especially via the very-real extraction and energy-production that feeds the massive computation involved in supporting such technologies?
Erode was a chance for us to learn and to gain some valuable skills in this fascinating space. We got familiar with Ethereum, Metamask and the IPFS (the InterPlanetary File System, in case you thought crypto types lacked ambition!). We look forward to exploring more with this technology, especially the challenges around security and sustainability. For now, the process raised important questions for our team, and hopefully in the minds of our clients and collaborators, too.
