Ordinals are Fiat. And Inscriptions aren’t Rare.
Part I. Introduction
Back in the good ol’ days, when you earned 8% safe yields on Celsius and Solana was the next big thing because good ol’ boy SBF said so, you had to mint your NFTs on Ethereum or a similar smart contract chain. As a burnt-out sometime artist myself, I was initially intrigued by the prospect. Digital scarcity, but for art? Maybe I should get back into it, and the hype would ease the task of bringing attention to my work. Yet, some skepticism seemed warranted, because I took more than one minute to think about how such a thing could be possible. On Ethereum, your cost of gas is directly proportional to the amount of data you ask the network to compute. Aren’t images data-rich, and therefore won’t it be prohibitively costly to execute transactions involving them? Oh, yes, that’s why the first NFT’s were exceedingly low-res little thumbnails — as a high-res digital photographer, that’s not exactly my style, but I’ll grant the idea was at least interesting. So, naturally I wondered how it was possible that so many relatively high-res NFTs seemed to be proliferating on NFT marketplaces. In the beginning it was difficult to find an answer to this question (it’s almost like no one wanted to talk about it, I wonder why), but it eventually became clear that most NFTs were just pointers to some off-chain, centralized data-storage provider which hosted the actual image. No art on-chain. Everyone was minting hyperlinks, and no one seemed to care. How fiat. Insert your favorite quote about the destruction of the money driving speculation and gambling to maddening heights in Weimar Germany.
But now, in 2023, things are different! We can now inscribe images directly into the Bitcoin blockchain, and if “Ordinals theorists” are to be believed, trade ownership of these images in some meaningful sense. Yet, once again, skepticism is warranted, if you take more than one minute to ask how such a thing could be possible.
Part II. Ordinals are Fiat
Before we look at Ordinals, let’s just consider what a Bitcoin transaction is, from the perspective of the blockchain itself. Very roughly, in a Bitcoin transaction Bob uses his private key to sign over some of his bitcoins from an address he controls, to an address Alice controls. A transaction is simply a record of this process, which is broadcast to the node network, and after sufficient confirmations, becomes immutable. Neither Bob nor Alice can do anything to make that transaction record disappear from the ledger. They cannot change its position in the total sequence of transactions. They cannot duplicate it, because those coins have already been spent. Other than the fact it references their public addresses and the amounts of coin that were transacted, the transaction viewed as an on-chain object has nothing to do with anything that either Bob or Alice can actually control. A transaction can include supplemental data in a section of the transaction record called the “witness.” The witness data stays with the transaction. It does not exist in Bob’s wallet prior to the transaction — it comes into being when the transaction is constructed — and it does not travel to Alice’s wallet after the transaction. When we inscribe jpegs onto the Bitcoin blockchain, we are inscribing them into the witness portion of a transaction, which does not move. If you take nothing else from this, please just absorb this: there is no concept of on-chain “ownership” of a transaction, and hence there is no on-chain concept of ownership of a jpeg inscribed within a transaction. Why then do “Ordinal theorists” act as though they do in fact own inscriptions?
The name itself, “Ordinal Theory,” the practitioners of which style themselves “ordinal theorists,” attempts to shroud the enterprise with a veil of erudition and gravitas, which, upon closer inspection, simply disintegrates. Let us consult the ordinal handbook itself (And for a more technical description of the method, here).
“Satoshis are numbered in the order in which they’re mined, and transferred from transaction inputs to transaction outputs first-in-first-out. Both the numbering scheme and the transfer scheme rely on order, the numbering scheme on the order in which satoshis are mined, and the transfer scheme on the order of transaction inputs and outputs. Thus the name, ordinals.”
The handbook also points out that all the way back in 2012 Bitcoin users, including Charlie Lee, were already proposing similar schemes for numbering satoshis. The handbook takes this as evidence that Ordinal theory is more akin to a discovery, rather than an arbitrary invention:
“These independent inventions of ordinals indicate in some way that ordinals were discovered, or rediscovered, and not invented. The ordinals are an inevitability of the mathematics of Bitcoin, stemming not from their modern documentation, but from their ancient genesis. They are the culmination of a sequence of events set in motion with the mining of the first block, so many years ago.”
The flowery language here creates the impression of a significant development — a weighty epiphany — ah, the Ordinals were there all along. Allow me to restate the discovery in plainer language, so we may evaluate its significance: Numbers exist, and we can perform logical operations on them. Bitcoin has had numbers in it all along; therefore, we can perform logical operations on those numbers. Profound.
In fact, we can perform any arbitrarily complex set of logical operations we wish. We may yet “discover” infinitely many internally consistent algorithms for stipulating labels for slices of bitcoins and tracking those labels across arbitrarily complex sequences of transactions. Starting at the genesis block, counting up from 0.00000001, and positing a first-in-first-out rule for matching inputs to outputs may be the most intuitive one, but we need not stop here!
So, Ordinals are a method of, first, pretending that it makes sense for satoshis to have serial numbers, and second, stipulating a set of rules to put them in order so that we can count and assign those serial numbers. In other words, we number the satoshis by fiat.
Maxmoney’s summary is helpful (the full article is recommended):
“Of course, this is nonsense. It makes no sense to say that “Output 8 came from the 3rd satoshi of Input 2”. Because inputs (less mining fees) always equals outputs, it is possible to do this “lining up” process for every transaction, but it does not contain any actual information regarding which satoshis went from where or to whom. To figure out the “ordinal number” of any “virtual satoshi”, start with any UTXO. Work backwards through all of the transaction history to line it up with previous outputs until you get to the coinbase in which it was mined. Add up all of the satoshis that had been mined until that point in time, and that’s the ordinal number. It’s a totally meaningless metric.”
The crucial point is that at the base layer, the Bitcoin protocol does not recognize this ordering. This counting method is imposed upon it by the third-party software that is your ordinal wallet running the algorithm. As Maxmoney (and many others) have observed, in reality a bitcoin transaction is a little bit like melting down some number of gold bars, and recasting them into new bars of differing sizes. The ordinal scheme is as if you imposed an imaginary three dimensional grid on your gold bars to subdivide them into arbitrarily numbered cubes before melting them down, and a similar grid on your new bars. Then you say “Grid cube (x,y,z) of bar X, pre-melt, corresponds to grid cube (x,y,z) of bar Y, post-melt. You can say such things, if you wish, but saying them and expecting to be taken seriously is foolish, to put it gently. To their credit, competent Ordinal theorists, including the original developer, are quick to admit this (well, they don’t admit that it’s foolish, but they admit the basic facts of the matter), but apparently, it bears repeating, because it fatally undermines the value proposition of “owning” a so-called digital artifact on Bitcoin. More exactly, it undermines that value proposition by undermining the notion that Ordinals and especially inscriptions are things that even can be owned.
This finally brings us to inscriptions. How is it that content is said to be inscribed “on” a particular satoshi? Again, from the Handbook:
“Satoshis can be inscribed with arbitrary content, creating Bitcoin-native digital artifacts. Inscribing is done by sending the satoshi to be inscribed in a transaction that reveals the inscription content on-chain. This content is then inextricably linked to that satoshi, turning it into an immutable digital artifact that can be tracked, transferred, hoarded, bought, sold, lost, and rediscovered.”
Or here:
“Inscriptions inscribe sats with arbitrary content, creating bitcoin-native digital artifacts, more commonly known as NFTs… These inscribed sats can then be transferred using bitcoin transactions, sent to bitcoin addresses, and held in bitcoin UTXOs.”
“Individual sats can be inscribed with arbitrary content, creating Bitcoin-native digital artifacts…”
Again and again the Handbook makes the bald assertion that sats themselves can be inscribed. Yet, as you delve into the technical details, the language starts to get a bit cagey. It appears to simultaneously acknowledge the reality that inscription content lives inside an indelible transaction, while maintaining the unsupported assertion that it is actually inscribed “upon” a particular sat.
“Additionally, inscriptions are included in transactions, so the larger the content, the higher the fee that the inscription transaction must pay.”
“Inscription content is entirely on-chain, stored in taproot script-path spend scripts. Taproot scripts have very few restrictions on their content, and additionally receive the witness discount [because this is where the content actually lives], making inscription content storage relatively economical.”
“The commit transaction commits to a tapscript containing the contents of the inscription, and the reveal transaction spends from that tapscript, revealing the contents on chain and inscribing them on the first sat of the first output of the reveal transaction.”
“The inscription content is contained within the input of a reveal transaction, and the inscription is made on the first sat of its first output.”
Again we see the bald assertion that the content is inscribed on the first sat of the first output. The only problem is that this statement is literally false in every way. We already saw that where the content (a jpeg, poem, other file, whatever) resides is in the witness data of a transaction — the Ordinal Handbook is quite up front about this. And we know that satoshis do not carry content from wallet to wallet. And as we already saw, it is literally impossible to “own” a Bitcoin transaction — it is just a record of what happened. Why then does the Handbook continuously aver that satoshis do get inscribed, and that you can own them? The only sense in which the satoshi — more exactly, the UTXO which, by fiat, is said to contain the sat — is linked to the inscription is that they have some shared history, on account of appearing in the same transaction at some point in the past. Nothing in the ordinal logic requires inscriptions to be linked to the “first” sat. So why is it the first sat of the first output? Why not the second sat of the first output? Why not the last, or one of the other hundreds, thousands, or millions of other satoshis which were also present in that same transaction? The answer is, well, because we are pretending that it is so. In other words, we say a sat is inscribed by fiat.
I’ve scoured the Handbook for a clear explanation of precisely why an inscription is bound to a particular sat, and I’ve found nothing but the bald assertions presented above. I suppose the reason for this is that if you say it like it is, it’s harder to trick non-technical users into attributing value to your project.
Consider Tuur Demeester’s inquiry and incisive counterpoints here.
https://twitter.com/TuurDemeester/status/1662193207227084801
He likens ordinal & inscription ownership to being offered the opportunity to purchase title to the planet Mars. I would refine the analogy further. It’s like being presented with a star chart of the actual night sky, with a novel “constellation” arbitrarily drawn on top, and being offered the privilege to pay money to “own” this constellation and all the stars within. The blockchain is the night sky, and every bit as immutable. The historical transactions recorded therein are the stars, every bit as untouchable. The jpeg is the constellation, every bit as arbitrary. And the chart depicting where to locate the constellation is the ordinal in your wallet, and every bit as meaningless. If you ascribe meaningful value to this charade, you either don’t understand what you’re buying (a problem this essay seeks to correct), or are at best acting imprudently. “But free agents value it on muh free market!” Yes, and we may simultaneously acknowledge that all valuation is subjective while also recognizing that wisdom is not.
Now, to further drive a wedge through the already-tenuous conceptual link between an ordinal-labelled sat and its associated inscription, here is some top kek for you. The default search method for an inscription in the Ordinal wallet or on the Ordinal block explorer does not even reference the Ordinal number. Once more from the docs:
“To create an inscription with the contents of FILE, run:
<ord wallet inscribe — fee-rate FEE_RATE FILE>
Ord will output two transactions IDs, one for the commit transaction, and one for the reveal transaction, and the inscription ID. Inscription IDs are of the form TXIDiN, where TXID is the transaction ID of the reveal transaction, and N is the index of the inscription in the reveal transaction.”
“Once the reveal transaction has been mined, the inscription ID should be printed when you run:
<ord wallet inscriptions>
And when you visit the ordinals explorer at ordinals.com/inscription/INSCRIPTION_ID.”
Screenshot for posterity:
The unique identifier for the inscription is just the TXID with an integer (not the ordinal number) appended to it. It is laughable that the simplest way to identify an inscription does not even reference the so-called digital artifact that is said to bear the inscription. You certainly can input the ordinal number into the block explorer and find the relevant inscription, but you need not. Obviously, it is a simple design change to incorporate the ordinal number into the inscription ID, but to do so would be an acknowledgement that there is no real link, it’s all fake, and they’re trying to obfuscate this fact. You could also imbed the ordinal number within the inscription content, but this carries no more weight than the U.S. government inscribing on the blockchain the statement “The USD is hard money.” Fiat written on the blockchain is still fiat.
In view of the foregoing, what “inscribe on a satoshi” really means is, if you count numbers a certain way up to a particular UTXO, then work backwards through that UTXO’s transaction history, you will eventually find a transaction with this particular inscription in it. Ultimately, therefore, what you have in your Ordinal wallet is a kind of off-chain “pointer” to a location on chain which is not controlled (owned) by any private key. Whereas Ethereum NFT’s are largely on-chain pointers to off-chain content, Bitcoin NFT’s are essentially off-chain pointers to on-chain content. In neither case is the thing to which you have any meaningful claim of ownership an immutable digital asset. According to the Ordinals Handbook itself, “Digital artifacts can have owners.” Therefore, inscriptions are not digital artifacts, because they cannot have owners. The only digital artifacts on Bitcoin are bitcoins.
Part III. Inscriptions are not Rare
Bitcoin is not designed to make digital art scarce.
The notion that blockchain-enforced digital scarcity is a property that art can have stems from a misunderstanding of how digital scarcity emerges from a blockchain. Digital scarcity is a property that emerges from carefully considered protocol design. A blockchain, when sufficiently decentralized across enough independent nodes, enforces consensus rules. If those consensus rules are engineered in such a way as to yield digital scarcity, then that blockchain enforces digital scarcity. E.g., the Bitcoin consensus rules contain an algorithm which logically ensures that there will never be more than 21 million bitcoins:
Hence, given sufficient decentralization, the supply is credibly capped, and bitcoins are credibly scarce. The Bitcoin consensus rules do not permit the appending of certain ledger entries without the requisite private keys (namely, ledger entries involving someone else’s coins). Hence, the ownership or control of those ledger entries is provably inviolable — we might say permission to append to the ledger is scarce. In your word processor of choice, you may create new ledger entries to transfer my sats to you as much as you want, but if you try to commit those ledger edits to the chain without my private key, the nodes will laugh you out of the mempool, because you are not abiding by consensus rules. Scarcity as a property inheres in bitcoins, which are themselves abstractions from the state of the ledger. Scarcity is not a property of any bit of information within the ledger. Indeed, strictly speaking, every bit of information in the ledger is replicated across the entire network.
The consensus rules do in fact permit the inscription of arbitrary sequences of bytes (e.g. jpegs) of data — information — into the witness data of a transaction. What the Bitcoin consensus rules do not contain, is anything whatsoever to restrict the long-term supply of jpegs on chain, nor any restriction on the ability to reproduce them. In fact, you can inscribe the same arbitrary sequence of bytes of data into as many transactions as you wish. So can I. After you pay your egregious fee to mint your “rare” piece of “art,” I can copy-pasta the exact sequence of bytes, and duplicate your “rare” piece of “art” (I won’t, because I’m not a rube who enjoys wasting what is scarce to obtain what is abundant, but I could). But I might make it my Twitter profile pic. Much scarce. Very digital. Wow.
In what is becoming a pattern of making spurious claims, once again the Ordinals handbook contends that inscriptions on Bitcoin are scarcer than NFT’s on, e.g., Ethereum:
Inscriptions are scarcer.
Inscriptions require bitcoin to mint, transfer, and store. This seems like a downside on the surface, but the raison d’etre of digital artifacts is to be scarce and thus valuable.
Ethereum NFTs, on the other hand, can be minted in virtually unlimited qualities with a single transaction, making them inherently less scarce, and thus, potentially less valuable.
And yet, we already saw there is no supply cap to on-chain jpegs. The supply of on-chain jpegs may approach infinity at a slower pace than Ethereum NFT’s, but it is still conceptually infinite. It’s only been six months, and there are already half as many inscriptions as there are bitcoin that will ever be mined.
In ~120 years, when the final bitcoin has been mined, how many inscriptions will there be? So. Heckin. Scarce.
We’ve seen how you can mint as many copies of a jpeg on-chain as you want. What you cannot do, is broadcast duplicate sequences of bytes that constitute a valid bitcoin transaction. This constitutes a double-spend, and violates consensus rules. The information contained in a valid bitcoin transaction cannot be duplicated at any later time. The information contained in a jpeg can be. Bitcoin the protocol is designed to enforce the scarcity of one thing and one thing only: the supply of bitcoins.
If you wanted to make a given jpeg provably, digitally scarce, you would need to engineer something like a protocol with consensus rules that check the entire history of the chain to see if that particular sequence of bytes had ever been inscribed before — but, leaving aside the feasibility of such a project, this isn’t much use because I can still mint the same image, less one pixel, ushering in an impossible process of filter refinement to reject those images that are “too close” to an extant one. This is, of course, leaving aside the fact that I can copy-pasta anything that appears on my screen, and an off-chain copy exists on the artist’s personal hard drive. If we have to trust the artist to destroy off-chain copies, what is the point of a “trustless” blockchain? Even if the consensus rules did somehow limit duplicate images, this by itself confers no value to a work of art. The starving artist is a meme for a reason. Uniqueness is not a sufficient condition for value accretion. How many unique pieces of analog art have little or no value? (Spoiler: nearly all of them). Converting to a digital medium and paying a bunch of money for a data center operator to distribute the image to a network of independent storage hosting providers adds value….how, exactly?
Alternatively, suppose you devise a way to inscribe an encrypted sequence of bytes, such that the jpeg is not viewable without your particular private key. But then you don’t get to display your degen trollery to the public and chase clout. I suspect many of the current participants would be less keen to blow sats on memes they can’t publicly display to “own the maxis.” Perhaps, bona fide art collectooor that you are, you could integrate your private key into a physical display screen in your posh moon-manor, to show off your avant-garde collection to your pump’n’dump profiteering social circle…but then, why not just own analog art? Once again, if you ever display your encrypted “rare” NFT in a publicly accessible digital format, you’ve compromised the only quality that made it rare. Moreover, the private key is information that lives off chain. You’re retaining off-chain the only thing that makes the art visible, so you’re not enjoying the benefits of “digital scarcity” that putting art on the blockchain was meant to provide. The scarcity is being enforced by your control of the private key, not the encrypted data replicated across the nodes.
“But but but, there are digital images of famous artworks, and the originals are still considered valuable!” Yes, because obviously a digital image of a Picasso is not a Picasso. The Picasso is a physical artifact existing in meatspace with a particular history and provenance. A byte-for-byte copy of a digital image is… literally the same digital image. There are no meaningful properties by which to distinguish them. “But if, by fiat, we count sats from the genesis block via an entirely gerrymandered method, we can point to the location on a distributed network of hard drives where the “original” jpeg resides…something something provenance!” Uh…congratulations…I guess? You came up with a convoluted way of doing something we could already do by just, ya know, looking up the TXID…oh yeah, that’s basically what the Ordinals explorer already does. Also, it’s not the original because you have to trust the creator to delete the original.
If you want to support an artist, why not just send him some sats in exchange for a beautiful piece in meatspace? Heck, you can even ask him for a digital piece, and ask him to destroy the original. You now have more control and exclusivity than you would if you inscribed it in a public ledger, in which case you would still have to trust him to destroy the original. The only reason the enterprise of putting art on the blockchain appears to add value, when in reality it adds none whatsoever, is because we are in the middle of a high time preference, fiat-brained gold-rush — a fake gold-rush of affinity scams, in which the application of the bitcoin/blockchain/crypto monikers to an endeavor creates a veneer of bleeding-edge innovation, to extract maximum value from speculators desperate to escape the manipulations of central bankers and kleptocratic, power-mad bureaucracies. But the tool for that escape already exists. It’s just Bitcoin. Don’t overthink it.
Part IV. Objections
- “Uhm…ACKSHUALLY, the term fiat implies the power to coerce or enforce! No one is enforcing the Ordinal scheme or the accepted conventions of inscriptions on anyone!” OK reply guy, don’t be pedantic. Obviously I am being rhetorical for effect. The relevant aspect of fiaticity for this discussion is the sheer fakeitude of it all (gosh heck now he’s making up words). Fakeitude, if you need clarification, is the lack of correspondence to an underlying reality. The problem is that less technical users are easily misled by the confident assertions of a reality that does not exist to buy something they otherwise might not. While I believe individuals are solely responsible for their own financial destinies, that does not absolve you of taking advantage of their inadequate due diligence.
- “We’re creating an indelible monument to last for eternity!” No, what you’re doing is defacing an indelible monument with your grotesque monkeys, inane pixel art, and pumponomics scam tokens. And once more, to make sure I thoroughly beat the dead horse, your defacements are infinitely replicable by anyone of a mind to do so.
- “Money is a collective hallucination, and so is Bitcoin, so why can’t ordinals and inscriptions be so?” Money is not a hallucination. The emergence of some good as money is a shared realization that that good has the best monetary properties. It is a shared realization that it is the most salable good, as Saifedean (and the fathers of Austrian economics before him) likes to say. Monetary properties are grounded in reality. A money emerges as more and more people recognize this reality. A hallucination is a mistake. In that sense, Ordinals and inscriptions are a collective hallucination — a hallucination that something is there when it is not: a claim to ownership of any meaningful asset.
- “Heckin’ darn it! I paid good money for this here inscription! Who are you to tell me I don’t own it?” This is not really an objection, but I’m trying to tell you that you got taken. Don’t be easier to fool than to be convinced you have been fooled. Take the L and get on with your life.
