Q&A: Bitcoin Ordinals, Inscriptions, and Digital Artifacts

Since early 2023, the Bitcoin community has been abuzz with the introduction of “digital artifacts,” or essentially non-fungible tokens (NFTs) on Bitcoin. In this guide, we answer the most pressing questions about what Bitcoin NFTs are, how they work, what the on-chain data shows, and potential benefits and concerns.

What are Bitcoin NFTs, Digital Artifacts, Ordinals, and Inscriptions?

Part of the confusion on Bitcoin NFTs comes from the fact that there are multiple terms being used interchangeably, some of which are incorrect.

Ordinals

An ordinal (or ordinal number) is a number that designates a place in a series or set, such as first, second, third, etc. This would be different from a cardinal number that states how many there are of something (one, two, three, etc.).

Therefore, ordinal theory can be applied to the Bitcoin network, where the smallest individual units of a bitcoin, known as a satoshi or sat, are ordered and numbered based on the order in which they were mined.

For example, Bitcoin’s genesis block (the first Bitcoin block ever mined) minted the first 50 bitcoin, and since there are 100 million satoshis per bitcoin, the Genesis block created sats #1 through 5 billion.¹

With this lens, every satoshi can be labeled with a unique number, somewhat like a serial number. However, this was not an intention of the Bitcoin protocol and is not included in the code. Therefore, numbering satoshis doesn’t require a protocol or core upgrade. Instead, it’s merely a way of viewing sats through this lens, which is done with special wallet software independent of the Bitcoin core software.

Therefore, ordinals are somewhat arbitrary and “pretend” as users can choose whether or not to recognize or participate in this naming convention. The core Bitcoin network doesn’t care about ordinals. To the code, the individual sats all look the same. The only thing applying this numbering system is this real-life human perspective. As such, anyone can apply a different lens and name or order sats according to a different convention. However, because this numbering convention was the first to do this and it’s a relatively simple concept, we believe it will likely remain as the established convention.

Why are some ordinals "rare"?

Along with this ordering convention, the creator of this ordinal idea, Casey Rodamor, has also proposed a way of adding “rarity” to different sats based on when they were mined.² For example, the first sat of a block (which is created approximately every 10 minutes) is deemed “uncommon,” whereas the first sat of the first block in a new difficulty epoch (occurs approximately every two weeks) could be deemed “rare.”

Again, this is just one interpretation. Other people could place more value on specific sats for other reasons, such as those that have palindrome numbers, are prime numbers, or merely because they have a personal affinity for certain numbers.

Ordinals are akin to coin or paper bill collectors that place a higher value than the face value on the currency due to its minting date or serial number. It’s a way for normally-fungible sats to potentially have numismatic value.

What does this have to do with inscriptions and NFTs? Technically, nothing. Ordinal theory applied to Bitcoin could stand on its own and people could start collecting “rare” or “exotic” sats, but when combined with inscriptions, ordinals can potentially be treated as NFTs.

What are inscriptions?

Inscriptions are pieces of data that are inscribed or “engraved” on the Bitcoin blockchain itself. This could be text, an image, audio, or even mini-games — anything that can be viewed in a web browser. This data put onto the Bitcoin blockchain can’t be destroyed or removed.³

Inscribing text and other data on the Bitcoin blockchain is nothing new, from the beginning when Satoshi inscribed the now infamous “Chancellor on the brink of a second bailout for banks,”⁴ onto the Genesis block, to when users put up to 80 bytes of data in a field known as OP_RETURN. What's different now is these inscriptions can take up more data (more on this later) and can be linked to an individual satoshi or ordinal.

In other words, if a piece of data can be written on the blockchain and also ascribed to a specific, numbered satoshi, then that can turn that satoshi into what’s effectively a token containing data that can also be sold, purchased, traded, etc.

What’s the difference between Digital Artifacts and NFTs?

When these sats are ascribed to a specific piece of data, such as an image, they’re what most people would think of as an NFT — digital tokens that represent something (e.g. a piece of art) that uniquely identify ownership. However, ordinal creator Casey Rodamor prefers to use the term “digital artifacts” as NFT is somewhat of a polarizing word, especially within the Bitcoin community.

These digital artifacts look and behave similarly to NFTs popularized on Ethereum and other chains. However, there are a few key differences:

1. Inscriptions are actually “engraved” or stored on the Bitcoin blockchain itself, not in some other location or server. This is different from Ethereum NFTs, which are a unique token that then can use different reference pointers or hashes to link a token to the data or image, but the data itself isn’t internally stored on the Ethereum blockchain. We think this is a major differentiator as the Bitcoin inscriptions are self-contained and don’t require additional servers or services, and are therefore safer for users who don’t want to worry about where their NFT is stored.
2. Because of the above, there is no way for Bitcoin NFT creators to receive royalty fees, which some other NFT creators currently enjoy.
3. Because the Bitcoin code views all ordinals equally (just another sat that is fungible with all other sats), it’s up to wallet software to provide functionality for ordinals and inscriptions. These wallets will need to practice coin control to make sure someone doesn’t accidentally send a specific ordinal as a regular transaction or fee payment.

Additional Questions

Are ordinals or inscriptions a threat to Bitcoin’s core code?

On a technical level, it doesn’t appear that ordinals pose any kind of technical threat to the Bitcoin core code or network. Why? Because ordinals are simply a way of looking at sats and do not require any kind of code change. It’s completely opt-in and is only usable by using an external and specific wallet or block explorer. Inscriptions also don’t appear to be a threat as they follow the network rules like any other transaction and also do not require a code upgrade.

If inscriptions didn’t require a code upgrade, what do they have to do with the previous SegWit and Taproot upgrades?

Technically speaking, nothing. Digital artifacts are a kind of byproduct that was imagined later. As previously mentioned, inscribing arbitrary data on the blockchain isn’t new and could be done pre-SegWit and Taproot. What these upgrades did was make it possible to store not only more data (in terms of size), but also do so at a cheaper cost, paving the way for digital artifacts.

SegWit (short for Segregated Witness) was a soft-fork upgrade activated in 2017. SegWit provided a number of benefits, including moving or segregating the signatures field of a transaction to a newly-created part of the block known as the witness field (hence the name).

This change effectively increased the block size limit from 1MB (or megabyte) to 4MB. The original block size limit was kept at 1MB, but the new witness field doesn’t count toward that limit. Not only can more information be packed into transactions and blocks, but the data put into the witness field gets an effective discount when it comes to fees since this data isn’t taking up the more valuable 1MB transaction space.

Taproot was another soft-fork upgrade activated in late 2021. The main Taproot upgrade removed the data limit per transaction. Therefore, it’s possible for only one transaction to take up the entire 4MB space in a block, something we saw in March 2023 when the biggest block ever mined.⁵

Does storing arbitrary data on the blockchain hurt Bitcoin’s usability, such as speed?

On a technical level, not really. The network continues to operate and function normally as this doesn’t affect the code. What’s happening is larger blocks are being created, which does incrementally increase the growth rate of the blockchain size, leading some to point out concerns of blockchain “bloat.” However, this doesn’t affect block times or transaction processing times.

In fact, verifying nodes don’t need to validate data in the witness section, but the data does need to be downloaded, so hard drive storage needs could increase incrementally. Others point out that this isn’t a concern because the total possible blockchain size isn’t increased (it will still be limited to 4MB per block) and storage needs remain small with hard drive capacity and declining costs far outpacing any additional incremental demand.⁶ In fact, only 40% of blockspace was being used prior to inscriptions. Now, this is closer to 80% as inscriptions are filling up unused blockspace.

Inscriptions could be a benefit for the Bitcoin network as it may increase the network’s overall security budget. If inscriptions are here to stay, this could increase fees and consistency of fees for Bitcoin miners, increasing mining desirability, thereby increasing the network’s security.

How does this affect Bitcoin miners?

Inscriptions and ordinals are generally a positive for Bitcoin miners, primarily because users are paying fees to miners to inscribe sats. (The inscription process is done through transactions and, therefore, inscription creators must pay transaction fees.) So far, over $4 million in fees have been paid from inscribers to miners.⁷ However, miner revenue from fees as a percentage of total revenue remains in the low single digits at 2% to 3%, although there was a spike to 5% when inscriptions started to gain adoption. It appears inscribers are relatively fee-sensitive and only willing to pay lower fee amounts because they don‘t have as much urgency to complete their transaction.

Miners also benefit from fuller blocks. For miners, finding a block has the same difficulty (and, therefore, the same electricity and computing power cost) regardless of whether the block is full or not. So, it's better to have a full block that generates more fees and maximizes the reward for finding that block.

There are also potential secondary effects that could impact miners if inscriptions are indeed here to stay. First, if inscriptions lead to consistently fuller blocks, this could increase miners’ incentive and provide a more consistent revenue stream of transaction fees, which are currently only a few percentage points of total miner revenue.

In addition, miners may be able to take advantage of entirely new revenue opportunities, such as selling rare or exotic sats to ordinal collectors, or even MEV-like opportunities, where additional revenue can be gained by efficiently building blocks.

Both of these could be important to the network and miners as the four-year halving cycle will continually decrease the block subsidy.

Won’t digital artifacts impact the fee market and crowd out “legitimate” transactions?

It's true that blockspace is a scarce resource and users wanting to transact on the Bitcoin network must compete with others. Therefore, inscriptions could increase demand for blockspace and increase fees for all users.

From a pure market perspective, the blockspace market is a free market, and since inscriptions are following the network’s rules and paying the fees, then the current fee is reflecting the supply and demand for blockspace as intended. In other words, inscriptions are just as “legitimate” of a transaction as anything else on the Bitcoin network.

However, while this is technically true from an economic perspective, it doesn’t mean it's a desirable outcome for furthering Bitcoin adoption. In other words, wealthier users may be driving up fees to create and trade images, while less wealthy users (such as those in developing countries) or those who want to complete “economic” transactions (such as making payments or transferring value) will now be forced to pay higher fees, eroding usability and adoption. This argument is somewhat bolstered by the fact that inscriptions are stored in the witness data field and only counts as one-fourth of the data “weight” as regular transaction data, meaning inscribers are effectively paying a lower fee than other users.

Overall, we think it's still too early to tell if either side of this debate is completely correct. We can look empirically as to whether fees are rising due to inscriptions. So far, it doesn’t appear to be alarming as median fees haven’t moved up on average with the jump in mempool activity or mean block size. What appears to have happened so far is inscriptions have lifted the lowest priority “fee floor” from 1 sat/vbyte, but there's limited evidence of “crowding out” at the higher priority fee level, according to an analysis by Glassnode.

Should the Bitcoin network be used this way?

We think both sides have legitimate claims and philosophical questions. Should the Bitcoin network only be used as a monetary system as originally set forth by Satoshi Nakamoto, or can it also be used for other purposes? This is more of a normative question, rather than a technical or engineering one.

As noted in the past, Fidelity Digital Assets’ Research believes Bitcoin is best understood as an emerging monetary network and monetary good. Therefore, we're wary of trying to use the Bitcoin network as a massive data or file storage system, as this was not what it was designed to do. Instead, we think there are better alternatives that were specifically engineered for functions like file storage.

That said, one of the Bitcoin network’s core features is its permissionless structure, which means some people may use it in ways others disagree with or didn’t intend and we'd be equally as wary of any proposed rule changes to try to stamp out uses that some subjectively view as bad.

In addition, while some worry about higher fees slowing Bitcoin adoption, things like digital artifacts or other future uses for ordinals and inscriptions may actually bring a different type of person to the Bitcoin ecosystem, thereby increasing adoption. These participants may initially come for the fun of “NFTs on Bitcoin,” but then end up discovering the other (and arguably more important, in our opinion) value propositions of Bitcoin.

Are digital artifacts here to stay?

So far, there's been quite a boom in Bitcoin digital artifacts with over 650,000 inscriptions created to-date.⁸ Will this continue, or will it be similar to how inscribing data using the OP_RETURN function initially had a burst of activity, but ultimately fizzled out? So far, the data shows that inscription creation has leveled off in the past two weeks, averaging around 10,000 to 20,000 inscriptions per day.⁹ However, fees paid by inscribers initially spiked in mid-February, declined, and then showed growth and another spike in late March.

We’re surprised at how quickly the ecosystem and infrastructure has grown alongside inscriptions, such as ordinals-as-a-service sites popping up, allowing anyone to create a digital artifact with no technical expertise (we tried one and it worked flawlessly). In addition, we're now seeing wallets provide support for keeping ordinals separate and even showing the inscription text or image in the wallet app itself. We're also seeing ordinal marketplaces develop.

Perhaps the biggest marker that “Bitcoin NFTs” are here to stay is the fact that Yuga Labs, the studio behind one of the most popular Ethereum NFT collections (Bored Apes Yacht Club), recently created a limited collection of 300 generative art pieces inscribed on the Bitcoin blockchain called “TwelveFold.” The auction for the pieces completed, netting a total of $16.5 million with the highest bid of 7.12 bitcoin and the lowest winning bid of 2.25 bitcoin.¹⁰

What could this mean for Bitcoin’s future?

First, we don’t yet know if digital artifacts on Bitcoin have staying power. Given the market for NFTs is still large, albeit lower than the 2021-2022 frenzy, we think it makes sense that a portion will move to and develop on Bitcoin given the self-contained nature of inscriptions where they live on the Bitcoin blockchain itself.

Second, we think many in the Bitcoin community may be more wary of future soft fork upgrades. While it doesn’t appear that these digital artifacts pose any threat to the Bitcoin ecosystem, both the SegWit and Taproot upgrades greatly and unwittingly enabled this. We think this proves the point that soft forks that look relatively safe may enable unknown capabilities that may then not be discovered until later. It’s not out of the realm of possibility that something less benign than NFTs on Bitcoin could be uncovered in the future that could require an additional or subsequent upgrade to rectify.

Third, while we don’t think storing arbitrary data on the Bitcoin blockchain is the network’s primary or even secondary purpose, we also don’t see the harm at this point. We’ll be closely monitoring the data to see if arguments of “crowding out” economic transactions do happen or if other negative side effects appear.

On the other hand, we can’t help but be somewhat excited for a new and rather fun feature that has been added to Bitcoin. We might even go so far as to say the fact that the Bitcoin network is now also an immutable, global “library” is a magical idea.

Finally, we think there’s a good chance this is only the first step or experimentation that could lead to other and perhaps more “practical” applications that additional data on the Bitcoin blockchain could later bring. Just like how the first internet webcam was a live feed of a coffeepot, set up to keep people from making an unnecessary trip and finding an empty carafe,¹¹ so too could ordinals and inscriptions be the first proof of concept for industry-changing Bitcoin capabilities and applications in the future.