DIDs and all textual IDs (e.g. ENS) cannot enclose a digital identity on its own. This is an immanent disadvantage of linguistic signs. Not only social relationships (such as SBT) but also visual representation (e.g. PFP NFT) are needed to build a digital-native identity. In this article, representation is given a new name as 'cyber flesh'.

In the physical world, the birth of an identity starts first with a body born, then a name would be given to this newborn baby. As a result, we will call the baby by this name throughout his/her lifetime. A correspondence is thus built between the name and the body, the signifier and the signified.

When we see our friends‘ and families’ names as unique. it is not because the name, the composition of certain letters itself is different, it is rather because the person, his/her face, body, and gesture are unique. Just like that no two leaves are alike in nature, the uniqueness of an identity is deeply rooted in the corporeal matter in physical life, rather than the linguistic sign we speak and abstract in our mind.

But now in the online virtuality, the birth procedure is significantly reversed: name first, then comes a body. No matter whether in Facebook, YouTube or Minecraft, one would first sign up with a non-duplicated name (e.g. ID, account, address) and by this registration, a default tabula rasa is instantiated immediately. Then people would add all the fruitful details to this blank canvas, including profile photos, outfits, banners, user-created contents (UGCs), etc. to make this default entity into a distinguished living, growing body.

This reversal of the birth procedure in the virtual world would result in a big problem, in that registering a name could only guarantee a default, non-duplicated entity but not a unique living identity. In other words, a valid DID could only confirm a valid account but could not tell whether the account belongs to a zombie or a human being. It may be influenced by coding language, in which a similar disturbing statement is framed as ‘a null pointer is a valid pointer’. But translating it into the ontological field of identity, we could only find some nonsense words saying zombie is valid human, or robots are valid users. Null is Yes, and that is why Sybil Attack remains a problem and a digital-native identity is so hard to born.

BrightID may be mentioned as a positive case to solve the problem above. By employing Zoom meetings to check every person behind BrightID registered, a valid name is therefore connected with a valid human being. Case solved. But I would argue that this Zoom checking section is hard to scale, and this checking is also highly centralised, with no difference from a central authority issuing an official ID. Additionally, the identity it confirms is still dependent and secondary to the identity in the physical world, making it a digital parasite rather than the ideal digital twin.

Though we are far from finding a favourable way to make all happy, this reversal of the birth procedure presumably has already lit up some novel sparks for us. For example, it may end a long-existing bias that favours words over imagery, thoughts over representations, the intangible over the tangible, and the metaphysical over the physical. Cognitive trajectory (such as password recovery question: ‘name a favourite childhood book’) or social relationships (your reference letter) are always heavily trusted when checking credits or recovering accounts. However, we could expect that in one day, digital hairstyles, avatar gaits, appearances of one’s e-cat and e-dog, and metaverse home decorations can all be treated as unique beacons to tell and distinguish a human’s soul.

We have long respected the abstractness and profoundness of words and thoughts, but maybe those ‘superficial’ pixel vignettes, audio clips, and digital models can make up the cyber flesh that we inherently lose in the digital virtuality. To support this argument, I will discuss further on ingredients of making an identity in the following section.

In the above section, I talk about the reversed birth procedure and the resulted shortcomings of Sybil Attack. Many have put forwards various ideas of identity prototypes to solve the problem and build a more robust digital identity. In many cases, social relationship data are heavily employed. Specifically, once a network of sociality can be portrayed, an identity can be anchored, located, and recognisable. One famous example of this kind is Soulbound Tokens (SBTs) from Weyl, Ohlhaver and Buterin’s paper, in which Soul, an on-chain identity prototype, is built by collecting SBTs, no matter credentials, proof of affiliations, or commitments, which are issued by other persons or organisations.

Heavily using social relationships to locate an identity seems sound and solid, but it casts a shadow already[1]. Because SBTs all come from extroverted elements, i.e. interactions with other people and engagement in a sociable situation, a soul is forced to always face outward because otherwise, the identity would risk being unrecognisable. Relations prioritise and outweigh the existence, meaning digital humans are defined more as nodes in a network rather than full and abundant beings. In this case, no individual boundary can grow up and no privacy or intimacy can be kept for this identity.

Back to our physical world where the identity notation is maturely developed and each person’s self can happily exist, we can always see a dual structure of identity: a person always consists of ex-vivo history and in-vivo character.[2] If the space exterior to a person’s skin is considered as public where s/he talks, plays, and interacts with others, then his/her eyes, fingerprints, souring throat, and imagining brain, are all safely situated in the private regime, namely the body, so that they are independent and cannot be easily changed by others’ words. So we can see that naturally people use both CVs as well as outfits to build up their self ‘image’ and show their uniqueness.

Therefore, here I propose that a digital native identity should comprise two parts: relationships as well as representations. Examples of representations include profile photos (PFP), avatar appearances and fashions, gaits, voices and sounds (e.g. VTuber’s voice and speech patterns), home pages, and so on. I give these representations for identity purpose a general name, cyber flesh, for their function to compose a cyber body in the virtual world.

The birth of ERC-721, a non-fungible token standard, has significantly expanded blockchain’s territory, and invited artists and creators alike into the crypto world where once settled only by investors, brokers, and engineers, etc.. NFTs’ non-fungible nature, meaning every token is different once minted, perfectly fits the need to create distinguished and unparalleled cyber flesh to make up a unique digital identity. Imagine not only profile photo NFT (PFP NFT) to work as someone’s face, but also voice-changer NFT to define a virtual idol’s speech sound, and fashion NFTs to make up a crypto KOL’s signature style.

Based on the advantages of non-fungible tokens (NFT) listed below…:

1. It is unique (non-fungible) in content.

2. It can be traded on blockchain in the universal form of tokens. Specifically, its homogeneous on-chain address, which works as the container to hold the idiosyncratic content, allows high liquidity in the market, and potential composability. (See my other article for further reading.)

3. All transactions are publicly transparent. This is benefitted from the overall infrastructure of blockchain.

…thus using NFTs as cyber flesh would be benefitted in similar ways:

1. Non-fungible contents sharing a common theme create culture memes de facto, therefore providing space for community with a consensus to grow upon. For example, Mfers, an NFT collection that consists of matchstick figure in a teasing gesture, attracts and hold attention with many collab and derivative projects from twitter to even offline space. A web2 predecessor would be Miku of Vocaloid, a virtual singer voicebank that inspires thousands of people to create performance video of this figure, and therefore build a large and influential community.

2. Tokenised transaction enables interchangeability and composability between different games or metaverse services while keeping the same identity. Imagine that a character I built in Decentraland can be translated into an XCOPY PFP or a Doodle PFP, keeping the face and clothes that I collect while rendering them into graffiti or mellow style. Or I combine a Loot bag of gears and a goblin PFP into Sandbox to create a goblin adventurer character that is actually playable in a game.

3. Transparency of records means all transitions are kept as part of a continuous and convincing history of ownership that everyone can access. Therefore, a digital identity can gain endorsement of provenance by making use of this self-certified history. Imagine a Captcha task in a metaverse service asking you to pick up hairstyles that your avatar has ever kept for the longest time, or one asking you to choose a more favorable home decor style out of several to check if it is consistent with your history decoration.

On the other side, it is true that using NFTs to build an identity that is so intimate to us might be also vulnerable to some disadvantages of NFT in general. For example, in most NFT collections, the original creators in fact keep the exclusive right of modifying or copying the art without the permission of its owner, unless they give certain licenses to clarify the rights as Yuga Labs did. So digital identities might face the challenge that the pair of angel wings one wears might be modified into Halloween bat wings all of a sudden, if no legal agreement is reached during purchase (but still it is a problem to all NFTs in general).

Another potential risk: given that a digital identity is usually a composition of a bunch of NFTs from different collections, can this visually integrated whole be regarded as a new creation independent of its miscellaneous components? And can the identity holder, the creator of this ‘meta figure’, use this integrated figure to do commercial activities and make money?

Therefore, I am looking forward to more NFT projects that come with more alternative practices on licenses and agreements. I think fewer exclusive rights and more legal attempts would expand the possible usage scenario of non-fungible tokens to make the ecosystem better.

This article mainly talks about three things: first, an analysis of the reversed birth procedure in a virtual world, and as a result, Sybil attack becomes a repeated problem. Second, a dual structure of identity is proposed in response to the birth reversal, specifically, ex-vivo relationships as well as in-vivo representations make up a full identity. And lastly, current advantages and disadvantages of NFTs as ‘cyber flesh’, i.e. digital identity representations, are checked with multiple examples.

I intentionally avoid to create a new identity prototype other than wallets that we already have, I also did not give any vague prototype names such as ‘Soul’, ‘Body’, or ‘Cyber human being’. This is not because I think a wallet is good enough to function as an identity. (In fact, the highly transferrable private key to a wallet is especially tricky for identity construction and an independent prototype for identity is needed.) It is just because I have not researched enough to write down this topic, and I suppose this part requires more words and deeper thoughts than what this article could present at the moment. This is a legit journey;-)

[1] Though in the DeSoc paper, SBT is introduced only to represent social identity to aid financial transactions, but misuse and overuse is inevitable meaning SBT would easily reapply to further areas other than economics.

[2] BAYC strategically parallel this dual structure with their business model, where people can use monkey face as profile photo to build (in-vivo) personal image on twitter, as well as join in the BAYC community to make (ex-vivo) social and business connections.

If you have any thoughts that you would like to share with me, please don’t hesitate to DM me on Twitter. Thank you for your reading!

DIDs and all textual IDs (e.g. ENS) cannot enclose a digital identity on its own. This is an immanent disadvantage of linguistic signs. Not only social relationships (such as SBT) but also visual representation (e.g. PFP NFT) are needed to build a digital-native identity. In this article, representation is given a new name as 'cyber flesh'.

In the physical world, the birth of an identity starts first with a body born, then a name would be given to this newborn baby. As a result, we will call the baby by this name throughout his/her lifetime. A correspondence is thus built between the name and the body, the signifier and the signified.

When we see our friends‘ and families’ names as unique. it is not because the name, the composition of certain letters itself is different, it is rather because the person, his/her face, body, and gesture are unique. Just like that no two leaves are alike in nature, the uniqueness of an identity is deeply rooted in the corporeal matter in physical life, rather than the linguistic sign we speak and abstract in our mind.

But now in the online virtuality, the birth procedure is significantly reversed: name first, then comes a body. No matter whether in Facebook, YouTube or Minecraft, one would first sign up with a non-duplicated name (e.g. ID, account, address) and by this registration, a default tabula rasa is instantiated immediately. Then people would add all the fruitful details to this blank canvas, including profile photos, outfits, banners, user-created contents (UGCs), etc. to make this default entity into a distinguished living, growing body.

This reversal of the birth procedure in the virtual world would result in a big problem, in that registering a name could only guarantee a default, non-duplicated entity but not a unique living identity. In other words, a valid DID could only confirm a valid account but could not tell whether the account belongs to a zombie or a human being. It may be influenced by coding language, in which a similar disturbing statement is framed as ‘a null pointer is a valid pointer’. But translating it into the ontological field of identity, we could only find some nonsense words saying zombie is valid human, or robots are valid users. Null is Yes, and that is why Sybil Attack remains a problem and a digital-native identity is so hard to born.

BrightID may be mentioned as a positive case to solve the problem above. By employing Zoom meetings to check every person behind BrightID registered, a valid name is therefore connected with a valid human being. Case solved. But I would argue that this Zoom checking section is hard to scale, and this checking is also highly centralised, with no difference from a central authority issuing an official ID. Additionally, the identity it confirms is still dependent and secondary to the identity in the physical world, making it a digital parasite rather than the ideal digital twin.

Though we are far from finding a favourable way to make all happy, this reversal of the birth procedure presumably has already lit up some novel sparks for us. For example, it may end a long-existing bias that favours words over imagery, thoughts over representations, the intangible over the tangible, and the metaphysical over the physical. Cognitive trajectory (such as password recovery question: ‘name a favourite childhood book’) or social relationships (your reference letter) are always heavily trusted when checking credits or recovering accounts. However, we could expect that in one day, digital hairstyles, avatar gaits, appearances of one’s e-cat and e-dog, and metaverse home decorations can all be treated as unique beacons to tell and distinguish a human’s soul.

We have long respected the abstractness and profoundness of words and thoughts, but maybe those ‘superficial’ pixel vignettes, audio clips, and digital models can make up the cyber flesh that we inherently lose in the digital virtuality. To support this argument, I will discuss further on ingredients of making an identity in the following section.

In the above section, I talk about the reversed birth procedure and the resulted shortcomings of Sybil Attack. Many have put forwards various ideas of identity prototypes to solve the problem and build a more robust digital identity. In many cases, social relationship data are heavily employed. Specifically, once a network of sociality can be portrayed, an identity can be anchored, located, and recognisable. One famous example of this kind is Soulbound Tokens (SBTs) from Weyl, Ohlhaver and Buterin’s paper, in which Soul, an on-chain identity prototype, is built by collecting SBTs, no matter credentials, proof of affiliations, or commitments, which are issued by other persons or organisations.

Heavily using social relationships to locate an identity seems sound and solid, but it casts a shadow already[1]. Because SBTs all come from extroverted elements, i.e. interactions with other people and engagement in a sociable situation, a soul is forced to always face outward because otherwise, the identity would risk being unrecognisable. Relations prioritise and outweigh the existence, meaning digital humans are defined more as nodes in a network rather than full and abundant beings. In this case, no individual boundary can grow up and no privacy or intimacy can be kept for this identity.

Back to our physical world where the identity notation is maturely developed and each person’s self can happily exist, we can always see a dual structure of identity: a person always consists of ex-vivo history and in-vivo character.[2] If the space exterior to a person’s skin is considered as public where s/he talks, plays, and interacts with others, then his/her eyes, fingerprints, souring throat, and imagining brain, are all safely situated in the private regime, namely the body, so that they are independent and cannot be easily changed by others’ words. So we can see that naturally people use both CVs as well as outfits to build up their self ‘image’ and show their uniqueness.

Therefore, here I propose that a digital native identity should comprise two parts: relationships as well as representations. Examples of representations include profile photos (PFP), avatar appearances and fashions, gaits, voices and sounds (e.g. VTuber’s voice and speech patterns), home pages, and so on. I give these representations for identity purpose a general name, cyber flesh, for their function to compose a cyber body in the virtual world.

The birth of ERC-721, a non-fungible token standard, has significantly expanded blockchain’s territory, and invited artists and creators alike into the crypto world where once settled only by investors, brokers, and engineers, etc.. NFTs’ non-fungible nature, meaning every token is different once minted, perfectly fits the need to create distinguished and unparalleled cyber flesh to make up a unique digital identity. Imagine not only profile photo NFT (PFP NFT) to work as someone’s face, but also voice-changer NFT to define a virtual idol’s speech sound, and fashion NFTs to make up a crypto KOL’s signature style.

Based on the advantages of non-fungible tokens (NFT) listed below…:

1. It is unique (non-fungible) in content.

2. It can be traded on blockchain in the universal form of tokens. Specifically, its homogeneous on-chain address, which works as the container to hold the idiosyncratic content, allows high liquidity in the market, and potential composability. (See my other article for further reading.)

3. All transactions are publicly transparent. This is benefitted from the overall infrastructure of blockchain.

…thus using NFTs as cyber flesh would be benefitted in similar ways:

1. Non-fungible contents sharing a common theme create culture memes de facto, therefore providing space for community with a consensus to grow upon. For example, Mfers, an NFT collection that consists of matchstick figure in a teasing gesture, attracts and hold attention with many collab and derivative projects from twitter to even offline space. A web2 predecessor would be Miku of Vocaloid, a virtual singer voicebank that inspires thousands of people to create performance video of this figure, and therefore build a large and influential community.

2. Tokenised transaction enables interchangeability and composability between different games or metaverse services while keeping the same identity. Imagine that a character I built in Decentraland can be translated into an XCOPY PFP or a Doodle PFP, keeping the face and clothes that I collect while rendering them into graffiti or mellow style. Or I combine a Loot bag of gears and a goblin PFP into Sandbox to create a goblin adventurer character that is actually playable in a game.

3. Transparency of records means all transitions are kept as part of a continuous and convincing history of ownership that everyone can access. Therefore, a digital identity can gain endorsement of provenance by making use of this self-certified history. Imagine a Captcha task in a metaverse service asking you to pick up hairstyles that your avatar has ever kept for the longest time, or one asking you to choose a more favorable home decor style out of several to check if it is consistent with your history decoration.

On the other side, it is true that using NFTs to build an identity that is so intimate to us might be also vulnerable to some disadvantages of NFT in general. For example, in most NFT collections, the original creators in fact keep the exclusive right of modifying or copying the art without the permission of its owner, unless they give certain licenses to clarify the rights as Yuga Labs did. So digital identities might face the challenge that the pair of angel wings one wears might be modified into Halloween bat wings all of a sudden, if no legal agreement is reached during purchase (but still it is a problem to all NFTs in general).

Another potential risk: given that a digital identity is usually a composition of a bunch of NFTs from different collections, can this visually integrated whole be regarded as a new creation independent of its miscellaneous components? And can the identity holder, the creator of this ‘meta figure’, use this integrated figure to do commercial activities and make money?

Therefore, I am looking forward to more NFT projects that come with more alternative practices on licenses and agreements. I think fewer exclusive rights and more legal attempts would expand the possible usage scenario of non-fungible tokens to make the ecosystem better.

This article mainly talks about three things: first, an analysis of the reversed birth procedure in a virtual world, and as a result, Sybil attack becomes a repeated problem. Second, a dual structure of identity is proposed in response to the birth reversal, specifically, ex-vivo relationships as well as in-vivo representations make up a full identity. And lastly, current advantages and disadvantages of NFTs as ‘cyber flesh’, i.e. digital identity representations, are checked with multiple examples.

I intentionally avoid to create a new identity prototype other than wallets that we already have, I also did not give any vague prototype names such as ‘Soul’, ‘Body’, or ‘Cyber human being’. This is not because I think a wallet is good enough to function as an identity. (In fact, the highly transferrable private key to a wallet is especially tricky for identity construction and an independent prototype for identity is needed.) It is just because I have not researched enough to write down this topic, and I suppose this part requires more words and deeper thoughts than what this article could present at the moment. This is a legit journey;-)

[1] Though in the DeSoc paper, SBT is introduced only to represent social identity to aid financial transactions, but misuse and overuse is inevitable meaning SBT would easily reapply to further areas other than economics.

[2] BAYC strategically parallel this dual structure with their business model, where people can use monkey face as profile photo to build (in-vivo) personal image on twitter, as well as join in the BAYC community to make (ex-vivo) social and business connections.

If you have any thoughts that you would like to share with me, please don’t hesitate to DM me on Twitter. Thank you for your reading!