Robin Sloan
The Media Lab
February 2021

A coat check ticket,
a magic spell

The Mint, 1809, Thomas Rowlandson
The Mint, 1809, Thomas Rowlandson

A while back, a dig­i­tal acquain­tance of mine went to work for Zora, a com­pany engineering a cryp­to­graphic protocol — really, a mar­ket — for artists. More recently, Zora published a bun­dle of devel­oper documentation and a client library for this pro­to­col. It can­not hap­pen that an art-adjacent SDK is pub­lished and I do not poke at it, so poke I did.

Here’s what I learned.

This message was emailed to the Media Lab committee. The assumed audience is subscribers interested in art and technology who are also fairly confident programmers. (Here’s more about assumed audiences.)

I’m not going to explain the bl — ch —  (never write it out; it’s like speak­ing the name of Sauron) because I presume you already know about this kind of dis­trib­uted cryptographic ledger. What you might not know about — I didn’t — is a spe­cial kind of entry designed to stand apart. The tech­ni­cal term is “non-fungible token”, with the ugly acronym NFT. It’s just a dis­tin­guish­able object with unique char­ac­teristics. There’s noth­ing rad­i­cal about that — this newslet­ter is non-fungible, in that sense — but these cryp­to­graphic objects are tracked (and bought and sold) inside a sys­tem that is oth­er­wise currency-like, so their dis­tin­guisha­bil­ity becomes noteworthy.

Radioactive dol­lars in Scrooge McDuck’s money bin.

Zora aims itself at artists, and its protocol deals in cryp­to­graphic objects asso­ci­ated with media: images, sound files, chunks of text, whatever. First and foremost, the pro­to­col allows you to cre­ate those objects, in an oper­a­tion called “mint­ing”, which is a nice bit of language.

This is where you dis­cover the whole sys­tem is more atten­u­ated than you imagined, because it’s not like you’re “uploading your media into the bl — ch — ”. No no no — noth­ing of the sort. The object is just a reference. It as­so­ciates your iden­tity with the hash of the media (remembering that a hash is a numeric fingerprint, ~unique to any sequence of bytes) along with the media’s URL, where it can actu­ally be viewed, played, read, whatever.

I minted some­thing; as a result, I now “own” a tiny digital object that refers to my short story The Writer & the Witch. This oper­a­tion didn’t have any effect on my own­er­ship of, say, the copy­right to the actual work; it just brought this odd new entity, this amulet, this coat check ticket into the world (and into my dig­i­tal wallet).

To mint the story and pro­duce the object, I wrote a short JavaScript pro­gram that used Zora’s client library to

  1. connect an Ethereum account that I control,
  2. compute the hash of the story’s text, and
  3. submit a request, including the account’s address and the hash, to the Zora contract on the Ethereum bl — ch — 

Note: this submission costs money! Ethereum’s trans­ac­tion fees are called “gas”, and the price of gas is shock­ingly high. When I asked my script to esti­mate how much I’d pay for the sub­mis­sion, it kept show­ing me an enor­mous number; I was sure it was a bug, and spent eas­ily an hour try­ing to fig­ure it out before realizing, oh, no, that’s … actu­ally the number.

And, whew, when you’re tin­ker­ing in this domain, you’ve got to be careful, because your mis­takes can cost you money. It really changes the feel­ing of run­ning a script when you know it might be about to burn $20 out of your wallet.

In the end, it did burn sev­eral $20s out of my wallet, and when the mint­ing oper­a­tion succeeded, it was with a ~$100 surcharge. If read­ing that makes you feel a strong wave of “why bother”, I don’t blame you!

But it turns out even egre­gious fees can’t over­whelm the plea­sure of get­ting some­thing to work for the first time.

What else can you do with Zora? You can buy and sell, of course. The vibe is “cool inter­net art gallery”. You find an object you like, place a bid, and, if you win, your name goes in the ledger. You can keep the object forever, lux­u­ri­at­ing in its aura, or you can sell it again.

That’s where some inter­est­ing capa­bil­i­ties pop up:

Each object main­tains a per­fect provenance—a chain of own­er­ship back to its creator. That’s legit­i­mately use­ful; it auto­mates some­thing that can be onerous, even treacherous, in the ~analog~ context.

When it’s minted, an object can have a profit-sharing agreement burned into its soul. Let’s say I sell the object I minted to the influ­en­tial col­lec­tor Katerina Hash-Jones and, sim­ply by dint of her own­er­ship, its value skyrockets. (Kate­rina only buys the best.) She sells it for 10X what she paid, but/and because of the agree­ment embed­ded in the object, half of that price is shared with me. And so on for the next sale, and the next, forever. My share could be 10%, or 1%, or 99%.

I think this is Zora’s most provoca­tive capabil­ity; it feels genuinely new.

It’s important to say here that all of this market activity is tracked and verified with the same whirling, churning, million-monkey mechanism that powers every bl — ch —  of this kind: an orgy of energy consumption, running nonstop. I think anyone participating in this market, or thinking about it, needs to contend with that fact.

So, I had tin­kered around, burned a few $20s, and minted my object. I was feel­ing mostly per­plexed by the whole experience, but/and then I read about a pair of projects that turned me around a bit.

The stu­dio called Larva Labs has pro­duced two pio­neer­ing bl — ch —  art projects: Cryp­toP­unks and Autoglyphs, both of which intro­duce sets of these objects — large but finite, Pokémon-like — that can be bought and sold. Cryp­toP­unks is more popular; Auto­glyphs is lovelier. I enjoyed this pod­cast interview with the Larva Labs duo; every­one in the con­ver­sa­tion is engaged and excited but/and also prac­ti­cal and at least a lit­tle bit skeptical.

CryptoPunks are sort of the skele­ton key for under­stand­ing this whole weird mar­ket, because (1) they were the first project of this kind, (2) they’re legit­i­mately fun and charming, and (3) they are now worth a LOT of money.

Top sale prices for CryptoPunks
Top sale prices for CryptoPunks

Eyes: popped! Nose: bleeding! Humans: weird!

Again, I’ll remind you — this is so so crucial — when you buy CryptoPunk #2890, you are NOT buying an image of a little blue dude, as depicted above. Rather, you’re buy­ing an entry in a ledger that as­so­ciates your iden­tity — yours alone — with Cryp­toP­unk #2890, an image of a lit­tle blue dude. That’s it. That’s the deal.

I think the wild mar­ket for CryptoPunks makes a strong case for the agree­ments avail­able on Zora. Larva Labs ini­tially dis­trib­uted these objects for free! If the Cryp­toP­unks had been minted with a 1% profit share burned in, their cre­ators would be get­ting a stream of income from all these eye-popping sales. Not a bad thing to imagine. (If you want to feel like you’re going insane, read this very dili­gent arti­cle about Cryp­toP­unk valuation.)

Listen: if you look at this stuff head-on, with a cold alien gaze, it seems absurd. I told you I “own” the object representing my short story, but if you change one char­ac­ter in the text — add one comma — it will pro­duce a dif­fer­ent hash, and you can hap­pily “own” the object rep­re­sent­ing that version, even though it’s func­tion­ally identical. Only the fact that I minted the object myself makes it meaningful. It’s all very tenuous.

But then: if you look at art head-on, with a cold alien gaze, it also seems absurd and tenuous. As the tren­chant Abe Burmeis­ter wrote on Twitter,

anthropologically speak­ing there is near infi­nite evi­dence humans like own­ing scarce objects and also like to turn com­mon objects into socially con­structed rep­re­sen­ta­tions of value. Com­bin­ing the two seems pretty reasonable, hard part is get­ting peo­ple to agree on which objects

This is really 100% social; it’s about con­jur­ing a dream of own­er­ship, of value. The Crypto­Punks were, and are, a magic spell; I mean that in a basi­cally literal sense.

Will Zora be able to cast a spell of its own? Impos­si­ble to know.

I think all of this feels more nat­ural to peo­ple who are immersed in the fine art market, or the market for rare limited-edition sneakers, etc. (I am not!) These are markets in which scarcity-by-design is a huge part of the fun, and that’s not true for all, or even most, mar­kets for cre­ative work. (I’m think­ing of the mar­ket for, say, stream­ing TV shows, or sci­ence fic­tion novels, in which the fun, as well as the value, comes from abun­dance and shared enjoyment, not scarcity and sole own­er­ship.)

So, while some of the com­pa­nies build­ing these pro­to­cols and plat­forms make grand pro­nounce­ments about rein­vent­ing the eco­nom­ics of like, ALL CRE­ATIVE PRODUCTION, I think it’s more accu­rate to say they are estab­lish­ing a new kind of fine art mar­ket, one with some use­ful and provoca­tive new capa­bil­i­ties. (And/or maybe, by extension, they are propos­ing the “fine art-ification” of a larger swath of cre­ative production … which is a coher­ent goal, even if it’s not one I support.)

The whole thing is silly, but a lot of things are silly, and humans do them anyway, and derive great plea­sure in the doing. I’ll confess: I’m curi­ous about what could be made and mar­keted in this way! One of the most inter­est­ing things Zora has pub­lished is here in their FAQ:

Instead of cre­at­ing arti­fi­cial scarcity by sell­ing copies of dig­i­tal goods … we pro­pose an alternative: make one original openly acces­si­ble to every­one no mat­ter who owns it, and sell that orig­i­nal token over and over again. As your work becomes more popular, peo­ple who want to col­lect it can buy the media — first from you, then from each other. Each time the work is resold, you get a share of the sale price.

You might detect some res­o­nance there with the “unlocking the com­mons” model that I’ve used sev­eral times, most recently with an e-book novella in 2020. There is a sense here of going with the grain of the inter­net; let­ting the bits flow freely, copied and recopied, while their coat check ticket, their daemon, that tight lit­tle cryp­to­graphic knot, sits serene, accru­ing value. Maybe.

My ini­tial explo­ration is complete; impov­er­ished by the price of vir­tual gas, I’ll let these ques­tions sim­mer for a while.

A follow-up on the orgy of energy consumption. I ran the numbers and discovered that the exploration documented above produced approximately 390 kg of CO2, equivalent to burning a barrel of oil. Like just … black smoke pouring out of my laptop keyboard here. The former figure is from a nice (by which I mean: horrifying) carbon footprint calculator for Ethereum wallets, the latter from an equally nice (😱) carbon equivalency calculator from the EPA. You can quibble with the accuracy of the Ethereum calculator in particular — it’s a tricky thing to estimate — but even if it was off by a factor of ten, the result would still be shocking. I bought a carbon offset for 3X the amount I produced; it cost $10, and is literally the least a person can do.

I’m curious to know your reaction to all this. Do you find it absurd but interesting? Oppressively Black Mirror-ish? Straightforwardly exciting?

(I guess I’m also curious to know if it even makes sense. If anything/everything could have been clearer, do tell.)

Whatever you choose, Robin will see that this response is coming from . Please also sign it with your preferred name ✌️


An inter­est­ing note from a sub­scriber:

I don’t make or sell art online, so this is just a theory, but one of the advan­tages of the cen­tral­ized plat­forms seems to be democ­ra­ti­za­tion of the abil­ity to pro­mote yourself. Value comes from the amount of atten­tion and inter­est (which I want to say is true for any piece of art, com­bined with the scarcity factor), so it’s an inter­est­ing thought exper­i­ment to try to guess how the creator-consumer ecosys­tem would adapt on a platformless sys­tem.

My intu­ition is that it will be very impor­tant for Zora and its peers to act as central gal­leries and mar­ketplaces. So, per­haps we’ll find our­selves in another sit­u­a­tion (there have been many) where the decen­tralized promise of the bl — ch —  lures peo­ple onto a some­what more traditional, cen­tral­ized platform — albeit one with some inter­est­ing properties.

It feels like the dig­i­tal utopi­ans (I have been one; I might be still) learn this les­son over and over: that accursed “cen­tralization” often coin­cides with accessibility, usabil­ity, efficiency, good design — the list goes on.


This is a use­ful contribution from Chris Krycho, who wonders

what might hap­pen if we tried Zora … but with­out the “mining”? And from this won­der­ing emerges a dream of a research pro­gram for technologists: What would the imple­men­ta­tion of such tech­nolo­gies look like? What would their edges and lim­i­ta­tions be? What would they afford, and what would they not afford? How might they be well-adapted to us and our scale? How might they be built for mutual thriv­ing with the rest of the crea­tures with which we share God’s green earth?

Sent to the Media Lab committee in February 2021