How did me and X3r0ne end up spending about a year researching 1995’s Hackers and creating pixel art and metadata based on every character, scene, and prop? Did we have some idea that was meaningful to NFTs when we wanted to do this? I think we did. We could be wrong but we thought it was neat.
Just as I was beginning to understand NFTs in 2020, a $based personality said in passing, “metadata is an art form.” I’ve thought about that a lot since then: how do you make the information that labels and accompanies your art into an important and substantial part of the art itself? Can you use *just* metadata to make art – and if you do, what’s that look like? What is metadata meant to do – is it meant to enrich the art or simply help categorize it?
If all there is to the art is the metadata, it’s something more like a literary work I think. There’s poetry minted to eth, I believe. That doesn’t mean it has to be a familiar form, though – text is taking on lots of new aspects even in print. Take, for example, David Foster Wallace using extensive footnotes and endnotes and linking them together and using them to add interesting and surprising texture to the writing or offering the sorts of fictional creations that we usually reserve for lists and catalogues. What actually struck me a little bit was cases where people seem to get really excited about metadata for its own sake:
Collectible culture, sports, comic books, music, movies.
It’s been said to my face that I need to shut off the impulse to talk about the song we’re listening to by rattling off all the facts and references and alternate version information that I know. It’s been said also to “please not ruin a movie” by talking about abstruse details of the set or cast or props or what lines are improvised and which are from the script and – and – and none of these details are really the piece of art itself. What I’ve found about movies and music and collectibles is that people are simply fascinated with the minutia – and the more of it they can sink their teeth into the better. (For instance – why is a promotional UNICEF van one of the most expensive LEGO pieces in the world?)
It’s also struck me before – particularly with the Ryder Ripps BAYC litigation – one of Ryder’s points was asking why a Bored Ape would have a Prussian Battle Helm. It’s a pretty good question, frankly.
I don’t really think BAYC creators were actually neo-nazis, but they claim it’s a random choice (and not a nod to edgy internet culture) – a lot of NFTs are weighed down with chaotic meaningless content. If BAYC are being honest, the helmet is just part of the stupid tendency to take a basic concept like “hat” and fill that value with random content. That is, to come up with a category for metadata and then fill it with whatever feels to me like putting the cart before the horse. You can see this in lots of NFT projects: “Oh, we should have glasses – make some up.” The glasses may be in the signature style of the collection, but without knowing for sure – I imagine lots of artists just crank out a handful of designs for the sake of making options. I find it really hard to care about random accessories, regardless of how scarce or common they are. Why would I want an ape with a prussian battle helm? What could that possibly mean to me or to whatever lore is cobbled together to make sense of the collectibles? Since BAYC won their court case I think officially the helmet is a random choice, and that probably makes it dumb.
It could be that metadata as an art form is too easily leveraged for financial speculation to be taken seriously, but just incase it can stand as an aspect of art or an art form, I’ll come up with some rules. These are obviously pretty subjective, but I don’t think I’ve seen a really good explanation of what metadata is doing artistically for collectibles since things really got out of hand in 2021.
1. Good metadata should fit the content rather than demand random content be created.
Looking at the content, then – I’ll talk about the two PFP projects I’ve worked on. Based Ghouls was built by Quincredible, all I did was advise and throw together a few backgrounds. What we discussed a lot, though, was the rich and varied memetic body of “Based Money” lore that we were drawing from. Eventually the Based Memetic Archaeological Society was born out of it – but the point here is that we almost literally mined old chats, telegram repositories, twitter feeds and old youtube videos and everything we could find so that the trait values for the collection reflected things that had actually happened or made sense in Based history. You could have a ghoul that existed on the moon or a ghoul that existed in vaporwave purgatory. It made just as much sense in the source material to have a ghoul with Robert Leshner’s hair as it did to have a ghoul with a pink halo. The important thing was: the metadata made sense. It was part of a consistent body of source material.
2. Rarity has to balance obscure content with signature content.
I remember when I was a kid growing up and collecting this toy or that – we always had favorites. We had least favorites. I had some He-Men toys that were grabbed from garage sales and were pretty old and tarnished. I loved to death one figure that was hard to find that had actual fake moss all over his body – it was so weird to have one warrior that had a felt skin when all the other ones were smooth plastic.
I had a few dumb looking guys that just had kind of random hats or goggles or beards and mostly just felt like they wanted to do as little as possible to make the figure “not He-man” but close. They were boring. Even though you might have 3 or 4 different versions of He-Man, you wanted the main character, of course. So there’s this odd balance when you look at a collection and ask what a person would want. On the one hand, the collector is going to want what feels like the Greatest Hits – the best and most known figures or characters or scenes or whatever. On the other hand, the collector might be really interested in some rare or obscure characters, misprints and prototypes and rare alternate versions – who knows! You can’t have the main character of the show be a rare toy – that would just frustrate everyone. At the same time, if all the rarest pieces are just dumb mistakes or hard to find random alterations, that’s boring too. It’s difficult to explain just what makes a rarity scale feel right instead of arbitrary and enhance the collectibility rather than frustrate the collectors.
3. Metadata should, if possible, enrich the art or collectible rather than just categorize it.
This rule is a little trickier to explain. All art and collectibles have some metadata (not all have a .json file or whatever but you know what I mean) just by virtue of existing: who made it, what did they call it (if we know) and did they write an explanatory abstract or say something about it in an interview? Other times I think things that could be considered metadata would be things like context. I was joking with some friends that after watching Tar maybe you should add Mahler’s marital status as metadata for his Fifth Symphony. We do that, in reality – we think people sometimes only understand a piece of art if they know all kinds of data about it that help us understand and make sense of it. Sometimes art is immediate and visceral and needs no explanation, but usually it seems like it at least helps to have some idea how to frame it. I had a project that was moderately popular called Lazlo Lissitsky – and the metadata of the individual pieces titles were, in my opinion, about 33% of the art. They had ridiculous paperback and pamphlet and poster type names – really ordinary silly stuff. The thing is, though, the titles were not at all descriptive of the art. They were fully invented non-sequitur imaginary uses that the art would have been put to – in a sense, the metadata stood alongside the art as a pairing with it rather than simply describing it.
4. Rarity should be meaningful rather than just statistical.
Someone was pointing out on Twitter a little while ago that they had a technically rare NFT. The acrobatics they went through to explain why were headache-inducing, something like: “This is the only Crypto Punk with short hair that doesn’t have glasses combined with the alternate beard structure OVER a blue background.” That can be as true as you like but honestly it just feels like a quirk of the rendering engine. If you create rarity by bending over backwards to find unique combinations of things that shouldn’t really be rare in themselves, it’s one of those situations where being technically correct doesn’t feel like it’s important.
I was just goofing off on the GlitchForge Discord a while back (circa late 2022) with some friends and we were chatting about cult classic movies we loved – projects we imagined we’d do glitch art for. I’ve always had a soft spot for HACKERS from 1995. I imagine a lot of people who like that movie that are messing with crypto right now. Anyways, we were joking about doing a PFP series for it – and my friend x3r0ne and I half-seriously decided to make some stuff for it. As we got further into making the little pixel versions we realized this had some real possibilities. It had potential in our minds, anyways, because we just enjoyed doing it for its own sake. We enjoyed categorizing the scenes, clothing, characters, and props from the movie. The trivia page on IMDB and a dozen other sites goes on for pages. Months later we were doing intensive research trying to figure out where this place or that was, what this character was holding in a scene, finding obscure little details that thrilled us. It’s really embarrassing how late it was in the game that I found out about the crew at hackerscurator.com – I read a ton of nandemoguy’s writing and corrected metadata in places to match his (if there was a conflict, which was rare). The thing is, the people at hackerscurator.com aren’t invested in the information about the movie – the metadata – because they are just making speculative investments or trying to *create* speculative investments: they genuinely enjoy knowing and researching all this stuff. You can tell.
The metadata we actually created for the project has the full names of things as far as we could figure them out: who made it, what it’s called, when it was made. There’s details from the movie that you only see for a few frames of the film – there’s little references to lines from the movie or scenes. The whole point was to delight people who were already fans of the film – to leave it so that an enthusiast couldn’t come back and say “I wish they had added ____.” You get to a certain point where you’re just adding dumb things like “the toast they ate in the morning” and you realize those aren’t meaningful items of metadata and you leave them out. You can get to a point where you’re just adding noise. And yet, it’s the metadata that differentiates two off-grey pixels into being the signature ring Angelina Jolie wears in a breakout role from any other two grey pixels.
5. If it makes sense, metadata should be a scaffolding for culture and content.
The last rule is considering how these flash developed cultures and communities and substance are built. When I did my artblocks autoRAD drop with Tartaria Archivist – I had generally subsumed a lot of what could be differentiating metadata into broader categories. That is, I wanted the aesthetics to be the *main thing* people observed and valued, not the labeling. Once we launched, however, people were already announcing that they had different things, like “vibing sprinkles” background and being enthused they were a part of that niche group:
In other groups like the Based Ghouls I’ve seen a similar pattern: the red halo ghouls weren’t intrinsically significant – it was just a variation on the pink halo that was from the community lore. That got taken and run with though, and the red halos are one of the traits I consistently look for now – similarly, the “inner circle” hood was just a joke about some old memes, but it was made into a private discord for a while by community members. Sometimes good metadata can be there to be latched onto – but predicting or controlling that can be really difficult. In fact, given the nature of self-ownership people pride themselves on with these kinds of things, it’s almost something you have to leave possible instead of directing. I can never guess now what aspects of a project metadata are going to become meaningful.
I think there’s a ripe field of material in movies yet for metadata harvest – making collectibles with rich and satisfying information heavy metadata that is meaningful to the content and balanced scarcity. The scarcity makes sense in the same way that comics releases have smaller run alternate covers, in the same way toy lines have minor characters and short run mistakes. The data is exciting in the way that knowing who has an uncredited came in the music video you’re watching is fun just to know – with no bearing on the music. Metadata as an art form may still take shape for high art, but in my opinion, so far, it lends itself really well to low culture. High art may have loftier sentiments built into it, but sometimes low culture is just more fun.