#26 - Demystifying the Riddle of the NFT
The Open Research Directions for NFTs and Digital Scarcity
Stanford Blockchain Review
Volume 3, Article No. 6
📚 Author: Matt Stephenson, PhD
🌟 Technical Prerequisite: Low/Moderate
Introduction
When the NFT explosion happened in 2021, the NFT research community did not do a good job explaining them to the public. One problem was that few of us had done any actual research. It’s also fair to say that we were much too proud about knowing what “fungible” meant.
As a result we blew some important opportunities to communicate the phenomenon. In March of 2021, when NFTs were just starting to get covered by media outlets, I was profiled by Vox but mostly rambled about “biographical indexicality”. Then the next month The Atlantic ran a long article titled “What Critics don’t understand about NFTs”, which seemed to promise a sympathetic explanation, and from two Harvard academics no less! [1][2]
But the article was sadly incurious, and just painted NFTs as an incomprehensible “craze”, and one that was interesting only insofar as it reminds us of the fact that people often buy things “not because of any innate worth” to them, but simply because they “expect others to value it later.” This perspective on NFTs, along with the fortunately now-moot arguments about climate impact, became the dominant and essentially sole NFT critique up to the present day [3].
There was something strange about these criticisms. Certainly they were right to call out the obvious NFT speculation that was and is going on. I said the same thing myself when I was interviewed. But critics often seemed to want to insist that, not only was there an NFT bubble, but that NFTs themselves were completely valueless. This is a strange claim. It would be like reading an article on the Housing bubble that also wanted to insist that nobody liked houses. Or a history book covering the Tulip craze that also kept arguing that tulips are all ugly.
Which is all to say that these articles were making one correct and commendable point – that NFTs had formed a speculative bubble – but tacking on another extraordinary point that this bubble had formed around absolutely nothing. That nobody could possibly like any of these things.
The Dominant Narrative and the Missing Link
The dominant media narrative split people’s opinion and understanding of NFTs. From outside crypto people were led to imagine that NFTs were entirely a grift, and perhaps the most cynical of grifts since supposedly nobody believed in or liked them.
Not everyone outside of crypto was so dismissive though. Ezra Klein gave a lucid and strong case for them as an interview prompt: “the ability to have verifiable, digitally scarce goods… is simply a functionality the internet needs. We have it in the physical world and it’s come into the internet”, and the great Stephen Wolfram liked NFTs and wanted to launch a collection. But the overall drum beat of “NFTs are a scam” grew and grew. [4][5]
And even within crypto the loudest voices often seemed confused, at least at first. One reason for this is that NFTs had been a genuinely grassroots phenomenon. They were then, and they are to this day, the only major application on Ethereum that was not predicted in the original Ethereum Whitepaper. Though NFTs were not completely new in 2021, they had been popular only within a small but serious community of artists, creators, and enthusiasts [6].
These OG NFT enthusiasts were not particularly prominent as thought leaders, and so most of the traditional crypto thought leaders were starting from scratch. The unfortunate result of this info-vacuum was an influx of interest in “Utility NFTs”, as if crypto-permissioning was best done by gas-inefficient tokens that featured irrelevant aesthetic metadata. Gating access with a utility NFT is like making your movie ticket a t-shirt or a poster. It might make sense (and has been done!), but only if people like t-shirts or posters, otherwise you’re just wasting your money creating an expensive ticket where a paper one (or an ERC-20 one) will do [7].
Furthermore, early opinions often talked around the fundamental question of whether people liked the NFTs themselves. Instead, NFTs were “a receipt”, “an autograph”, a token granting some special “access to the artist”, etc. Whatever the story, it seemed the NFT’s value was always elsewhere, in some intellectual property, some other valuable thing that was being autographed, recorded, or accessed by the NFT.
Whatever we think of these theories now, it appears that nobody bothered to actually research them. Were people collecting free floating artist signatures? Were they really buying an artist’s attention in a weird way? Were they really buying and selling receipts? All would honestly be pretty interesting.
These theories certainly did not make NFTs seem more understandable or appealing to outsiders (“come pay to talk to artists you’ve never heard of!”). And compounding the matter further was the fact that even prominent thinkers who did understand NFTs were not always crystal clear in their writing about them. For instance, a 2022 paper from Glen Weyl, Puja Ohlhaver and Vitalik Buterin noted that “blockchains offer traceability in transactions that prevents someone from right-click copy-and-pasting a valuable NFT (and sybil attacking the original owner)”. This is completely correct, but it would also be extremely confusing to an outsider–the technology does not “prevent” anyone from copy and pasting the image of course. Only when “the NFT” is understood to mean something like “a digital object uniquely tied to some set of events” does their point make sense [8].
By 2022 it might have seemed obvious that NFTs were being differentiated and valued based on their unique histories (rather than, say, their aesthetic qualities), and so Buterin et al. offered no citation for their claim. But suppose they had wanted to – what could they even cite? No research had been done.
A Personal Journey into NFT Research
My research into NFTs is mostly a bit of history – I was intrigued by NFTs during my PhD but, as an economist (even a behavioral one) it seemed pretty far afield. As far as I can tell, the phenomenon of valuing an object based on its unique history is completely commonplace in the real world and yet completely novel in the digital world until NFTs. That is, someone might proudly display in their house the real home run ball that Barry Bonds hit, but nobody seemed to ever ask which mp3 file is the “real one.”
This phenomenon is novel and important in the history of digital life. Just as privacy is something taken for granted in the real world–the ability to close a door and be truly alone for a bit–so too is uniqueness. And if privacy is fairly scarce in our digital lives, uniqueness was almost completely unheard of.
So as a side project, I looked into markets for goods which seemed to resemble NFTs. Art markets are weird and contrarian, so I settled on the more home-spun memorabilia market. The interesting example I found was James Naismith’s typewritten Rules of Basketball, which were auctioned off by Sotheby’s for $4.4m. Naismith’s two typed pages are now proudly displayed at Kansas University.
The Naismith example was so interesting to me since, being typewritten, the documents are not aesthetically unique. That is, if a bidder truly just wanted that text on paper from an 1890s typewriter, they could have saved millions by just buying Naismith’s model of typewriter for a few hundred dollars. A sort of retro “right click, save as”.
Sotheby’s promotional materials for the Naismith auction emphasized that the document was “typed up the very morning that Naismith introduced his new sport to the world”. This suggests that the appeal lies in some notion of continuity. And this continuity angle is an argument for why a blockchain specifically would be important for NFTs. That is, we couldn’t ask which mp3 was the “real one” because mp3 files didn’t have a continuous history in any relevant sense. We could now ask which art was the “real” art, because the blockchain could make sense of such an idea [9].
The Continuity Theory of NFTs
The Ship of Theseus Paradox is a famous example of our conflicting intuitions about identity. Theseus’ ship has been restored by replacing every single wooden part and then, in Thomas Hobbes’ version, the discarded pieces of the “true” ship are reassembled into another identical ship. The former ship, with replaced parts, is continuous with Theseus’ original ship, but the latter, with original parts, is now identically composed to the original. This paradox reveals a tension in our intuitions about which of two objects is the “real” one, here teasing apart continuity vs. composition.
Note, however, that the composition argument relies on continuity as well. In Hobbes’ version of the Theseus paradox, someone “had kept the Old Planks as they were taken out, and by putting them afterward together in the same order, had again made a Ship of them”. That is, the “re-composed” Theseus’ Ship has planks that have a continuous identity over time tracing back to the original Ship. If it didn’t, and the planks were new, we’d just call the ship a replica.[10] {1}.
Continuity thus seems to be a core feature of identity and, indeed, research on infant cognition by Elizabeth Spelke suggests that continuity is the “central’’ concept of identity. It requires an object to trace “exactly one connected path over space and time”. This is, of course, what the ERC-721 standard provides and is thus suggestive of why unique digital objects emerged with NFTs. [10]
A study of digital cognition could, for instance, repurpose some of Spelke’s experiments in a digital context to see if this notion of continuity makes a difference in people’s perception of digital objects, as in the outlined experiment in the figure below.
An experimental design contrasting (1) The UTXO model of identity, (2) The ERC-721 model, (3) The non-blockchain OSI model, the sidechain (NFT) model. In each panel, an object A that begins in period 1 on the left progresses through some change in period 2. Then in period 3 (on the right of each panel) we ask whether we can tell which object is A.
The continuity theory of NFTs received some supportive evidence with the rise of Ordinals/Inscriptions on the Bitcoin Blockchain. The Bitcoin UTXO model, as in the above figure in quadrant (1), does not enable continuous identity natively because transfers split the “object” without distinguishing between its parts. But Ordinals proposed a “first in, first out” interpretation of UTXO transactions which provided, at least at some level, continuity. With Ordinals, and a notion of continuity, people began to value Bitcoin-native NFTs {2}.
Open Research Directions and Questions
But that is only one theory, and it’s surely an incomplete one. So I will close this essay offering some other potentially fruitful avenues of research.
1. What technical aspects of NFTs serve to differentiate them, and how/why?
This research could be market-driven, evaluating whether for instance “on-chain NFTs”, which are reproducible entirely from the contracts and tokens themselves, matter to demand? Or what about NFTs which have their referenced media stored on IPFS or otherwise hashed in the metadata?
This research could also be experimental or psychological, abstracting these technical qualities into whatever assurances or signals they provide. Ideally, a theory of a “minimally viable NFT” could be developed, identifying the essential qualities an NFT should possess.
2. How “chain specific” are NFTs? Are bridged NFTs like Emblem Vault, treated as “real”? Why?
Bridged NFTs, which are NFTs which originate on one-chain but are owned, bought, and sold on another, represent a fascinating test case. Though Ethereum-native NFTs dominate the market, applications like Emblem Vault have “bridged” over valuable, often historical, NFTs from other blockchains. Emblem Vault alone has seen nearly $100 million dollars in transactions of these bridged assets.
Is it the blockchain architecture that creates this interoperability? Or, equivalent, could we predict what assets could not be bridged? Is there something quality enhancing about a bridged NFT? E.g. Is the NFT perceived to be more real, or more long lasting on a larger (or more decentralized) blockchain?
3. What is the nature of an “NFT collection”
A notable feature of the NFT world is that many are sold as parts of “collections” of aesthetically and/or conceptually similar items. One reason for this appears to be tradition, as the first prominent Profile Picture (“PFP”) collection, CryptoPunks, was an edition of 10,000. However, the market into which 10,000 CryptoPunks were sold was tiny compared with the market into which, say, the 10,000 Azuki PFPs were sold some 4 years later. Is tradition really so strong? Or is there something else important about this number?
One possibility is that in some contexts NFTs create a “minimal group” which enables greater cooperation. Owners of a Bored Ape would, in this analogy, be like the experimental participants who expressed a preference for Klee over Kandinsky in the classic experiments. What are the analogues for this sort of behavior in art, fashion, or other markets? [11]
4. NFTs and digital property rights
In property rights theory, ownership is the possession of some set of rights over a resource, most particularly (1) the right to exclude others from the resource ("exclusivity") and (2) the right to sell or transfer ownership of the resource ("alienability".) These rights may be enforced by a third party, such as legal enforcement by police officers acting as agents of a court, but they may also emerge as a result of non-legal cooperative action. [12]
NFTs are a striking example of non-legal cooperative enforcement via blockchains, with a clear provision of exclusivity and alienability. This makes them a natural fit for the study of non-legal property rights. There is a further overlap in more recent theories of property and the public domain. To take an example from Stephenson 2022, adapting an argument from Yoram Barzel, “suppose that you own a sandwich and I take a bite without permission. In that case I have violated your property rights. But if I look at the sandwich, or even photograph the sandwich and hang the photo on my wall, I likely have not violated your property rights. Thus the visual attributes of a displayed sandwich are in the public domain. In a similar way to the sandwich owner, NFT owners seem content to openly display the aesthetic qualities of their NFT, allowing anyone to view them, download them, etc. for free.” We might then ask what would constitute a “bite of the sandwich” for an NFT? [13]
These features of blockchain-based NFTs make them especially suitable for study in Property Rights Theory. Not only are they novel examples of effective non-government property enforcement at scale, they are also spartan about what lies in the private domain: essentially just an NFT’s tokenid and contract address pair are excludable and alienable.
5. NFTs, Metaverse, and Display
Metaverse pioneers proposed a vast VR world in which we could all play and build. It was originally imagined as a shared substrate owned by a single organization–the “Association for Computing Machinery”-- and the political economy of such a scenario is more than a little concerning. I’ve argued elsewhere that NFTs, as shared virtual objects, are essentially the metaverse arriving backwards; “object first” rather than substrate first.[14]
Existing NFTs largely colonize fragments of “metaverse” space we’re offered on the modern internet, such as the PFP spaces most social media allows. But what other spaces exist? And do these determine the way NFTs are used, consumed and valued?
Further, there is interest in exhibiting NFTs on digital displays like the Samsung Frame. What are the emerging norms around such a display–is there an expectation that display signals ownership? That is, would displaying an NFT that you don’t own be more comparable to displaying an art print (not deceptive) or wearing a counterfeit jacket from a popular brand (deceptive, at least in some contexts.)
Conclusion
Researchers have fertile ground ahead in investigating the technical, psychological, and sociological aspects of NFTs. The economic relationship to evolving notions of digital property rights, their role in the broader metaverse, and their implications on digital display and authenticity are all further topics ripe for exploration. Each of these areas not only promises insights into the mechanics and nuances of NFTs but also offers reflections on our evolving digital culture and the intertwining of technology with societal values.
The advent of NFTs has revealed a chasm in our collective understanding of digital ownership and value. The discourse has often been polarized, oscillating between the extremes of dismissiveness and unbridled enthusiasm. In such times we need research that can help unpack and explain the NFT phenomenon for what it really is. NFTs are surely less than what their most enthusiastic proponents suggest, but likely more, much more, than the critics suppose.
About the Author
Dr. Matt Stephenson is the Head of Cryptoeconomics at Pantera Capital. He was previously a Senior Behavioral Mechanism Designer & Engineer for BlockScience, working with organizations from Filecoin to MakerDAO. He's lectured at Columbia University on DAOs and NFTs; his doctoral thesis covered price discovery and endogenous value in crypto markets. Matt has a PhD in Behavioral Economics & Strategy from Columbia University, Master’s in Strategy from Columbia Business School, and M.S. in Economics from the University of Warwick as a Fulbright Scholar. He is from Montana.
Special thanks to Matt Condon and Mitchell F. Chan for their comments on this essay (and their insightful thoughts more generally).
References
[1] “The Value of NFTs, Explained by an Expert” on Vox.com https://www.vox.com/the-goods/22358262/value-of-nfts-behavioral-expert
[2] https://www.theatlantic.com/ideas/archive/2021/04/nfts-show-value-owning-unownable/618525/
[3] See artists calling NFTs a “Ecological Nightmare Pyramid Scheme” in 2021: https://www.theverge.com/2021/3/15/22328203/nft-cryptoart-ethereum-blockchain-climate-change
[4] https://www.nytimes.com/2022/04/05/podcasts/transcript-ezra-klein-interviews-dan-olson.html
[5] A conversation between Nassim Nicholas Taleb and Stephen Wolfram at the Wolfram Summer School 2021 (NFT conversation begins 1 hour and 25 minutes in.)
[6] https://cointelegraph.com/news/ethereum-white-paper-predicted-defi-but-missed-nfts-vitalik-buterin
[7] In 1990, Disney used t-shirts instead of tickets for admittance to their movie Dick Tracy. The t-shirts proved popular, adding $1.5m to the opening gross and based on a scan of ebay they still sell for 3-5 times the original cost. https://en.wikipedia.org/wiki/Dick_Tracy_(1990_film)#cite_note-43
[8] https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4105763
[10] https://psycnet.apa.org/record/1993-98597-005
[11] https://en.wikipedia.org/wiki/Minimal_group_paradigm
[12] https://www.google.com/books/edition/Ownership_Control_and_the_Firm/Ds4nyAEACAAJ?hl=en
[13] https://academiccommons.columbia.edu/doi/10.7916/ngtm-3a13
[14]
https://twitter.com/stephensonhmatt/status/1649475415273275392
Footnotes
{1} This does not reconcile the tension, but instead suggests that some notion of continuity is necessary (but not necessarily sufficient) in establishing an identity.
{2} Earlier experiments on Bitcoin like Counterparty are fascinating but more complicated and are not treated here.