From Decentralized Society: Finding Web3’s Soul:

Property rights are defined in the Roman legal tradition as bundles of rights to use (“usus”), consume or destroy (“abusus”), and profit (“fructus”). Rarely are all these rights jointly vested in the same owner. Apartment leases, for example, confer limited rights of use (“usus”) to the lessor, but not unfettered rights to destroy the apartment (“abusus”), sell it off (“fructus”), or even transfer use (subletting). Rights of real property (land) are typically encumbered by a range of restrictions on private use, grants of public rights of access, limits on rights of sale, and even rights of purchase by eminent domain. They are also typically encumbered with mortgages that transfer some financial value to lenders.

The future of property innovation is unlikely to build on wholly transferable private property so far imagined in web3. Rather innovation will hinge on the ability to decompose property rights to match features of existing property regimes, and code even richer elaborations.

Fees from token lending/renting for token-gated access are analogous to cash flows for usage-based software business models, thereby providing an intuitive valuation framework for tokens. The use of the term “token” rather than “NFT” was done intentionally for two reasons:

1. Token-gating mechanisms can accommodate fungible tokens, not just non-fungible tokens. In the same way that you can restrict access of a piece of content, Discord channel, livestream, etc. based on whether or not someone owns a particular NFT in their wallet, it’s also possible to restrict access to services based on whether this person owns 0.1% of the total supply of some ERC-20 token or whether this person has owned some quanta of an ERC-20 token for, let’s say, over a year.

2. The meaning of the term “NFT” varies wildly depending on who you ask to define it. Most people associate NFTs with pictures rather than the token standard that points to data that may or may not be an image — the medium is confused with the message. Furthermore, the emotional associations and connotations around “NFTs” often leads to critics and proponents talking past each other at most levels of discussion.

It is for these two reasons that, while I will primarily be referring to token-gating in the context of token-gating based on NFT ownership, I will use the term “token” instead of “NFTs” wherever I can.

But okay, here’s the point: token valuation methodologies in crypto, and especially around NFTs, are a vibrant area of contention due to the lack of cash flow streams (or, at least, the ability to “turn on” cash flows, which would be somewhat analogous to a SaaS company maturing through its company lifecycle deciding to de-emphasize growth and market capture in exchange for higher margins and FCF) that can be discounted back into the present for most cryptoassets. It’s argued that certain cryptoassets can be valued through the discounting of value flows back into the present because they are producing or will produce natural yields as a byproduct of staking (e.g., PoS Ethereum post-merge) or the protocols that the cryptoassets represent equity and governance ownership in either charges fees (or contains the ability to charge fees and alter take rates) for services rendered (this is essentially the argument for why governance tokens for DeFi protocols, like Uniswap, have value).

Although there exist NFT collections that enable holders to more NFTs through voting mechanisms or breeding (Nouns being an example of the former and Axie being an example of the latter), I don’t believe investors in these collections make valuation claims based on any DCF analogues. There’s perhaps an argument to be made that the BAYC collection and ApeCoin might represent intangible assets that maybe could be valued based off of some expectation of future merchandise, digital content, and event sales made by leveraging brand value [a business strategy exemplified by Disney’s leveraging of their characters and cinematic universes to drive content subscriptions and park tickets], but even this is an argument that represents a stretch for most traditionally-minded investors. For now, the most salient existing examples of NFTs that could be valued based on some expectation of future cash flows are those that correspond to virtual spaces (i.e., digital “land”) which can be rented by advertisers, thereby making them to analogous to traditional web2 advertising spaces (newsfeeds, articles, websites, search engine results, etc.) that can be valued using metrics like CPC, CPM, CPI, CPA, and CTR [some analysts have provisionally applied multiples-based real estate valuation frameworks around the similarity of the properties being assessed on digital “real estate” or digital “land” in (what I believe to be) an overly skeuomorphic manner].

All of the aforementioned valuation frameworks for NFTs I’ve just mentioned are speculative at best and would be readily dismissed (and probably ridiculed) by most investors who aren’t VCs given the prevailing attitude around all things NFT and Metaverse. I believe, however, that tokens that can be rented and used to access online services will undeniably have intrinsic value in a manner that clearly avoids the all-too-common self-reference that helps people justify speculation in the crypto space.

The most obvious example is that of a token-gated streaming service in which viewers subscribe through purchasing a token that corresponds to access to that day’s/week’s/month’s/year’s content — there are undeniable parallels between valuing the associated tokens of this hypothetical service to the valuation of media companies like Netflix and Spotify that traditional investors are already comfortable with. A skeptic following the argument up to this point might object “Why would a media organization choose to jump through all these hoops of token-gating their content and then issuing tokens, I don’t see the point” or “A media company utilizing these mechanisms will be outcompeted by centralized media services due to difficulty of customer acquisition as well as economies of scale in content production” or something along these lines — I must emphasize that the competitiveness of using token-gating mechanisms is a separate consideration from the recognization that valuation methods from software and emdia map well onto this hypothetical token-gated streaming service. Put even more concretely, if Netflix were to switch their subscription log-in and billing/payments systems architecture exclusively to a token-gating and token issuance model [they wouldn’t and they won’t for inumerable reasons, I’m just saying IF they did], then the value of the tokens involved would be undeniable.

Let’s make our example less obvious and less familiar. Let’s say that Netflix minted a limited supply of 10,000,000 access tokens and could rent these access tokens out for potential viewers of their content library [yes, they wouldn’t do either of these things, but please bear with me]. And let’s say that viewers could rent by the second such that they effectively pay-to-view on a second to second basis depending on how long they rent out the token that grants them access to Netflix’s content library. Therefore, someone who only watches one show during one month is charged significantly less than someone who binges two seasons and a movie every week of the same month simply — this would be tantamount to usage-based pricing, with the usage in this case directly corresponding to time spent (and therefore indirectly linked to costs from cloud compute, content caching, maybe even content amortization). This hypothetical, token-friendly Netflix service could charge different token rental prices based on the content being consumed (i.e., rental of tokens that can unlock access to more recent movies and shows cost more (a concept already very familiar media streaming services and media analysts)), the resolution quality of the content, and/or the ownership of other tokens in the viewer’s wallet (if the viewer’s wallet contains a token issued by a school that indicates that she is currently a student or by a government that indicates she is a citizen domiciling within a certain country (again, a concept already very familiar to streaming services)), etc. In this case, too, it is undeniable that each of the 10,000,000 hypothetical tokens could be valued intrinsically based on some expectation of future cash flows by Netflix’s internal finance team — why they would choose to issue 10,000,000 tokens in hypothetical might stem from a desire to moderate the variable costs from viewer demand on the necessary cloud computing and CDN services required to run personalization algorithms and stream content. But Netflix, like most internet services running at hyperscale on CSP infrastructure, already have predictive models for load balancing and pre-caching content in expectation of viewer demand, so, for this reason and others, the example I’m providing here is a stretch.

So let’s contort this example one last time, to make it less familiar and more crypto-native. Let’s say there was a DAO called “Ethflix” (pronounced “Eth”, not “Eth”) that operated a video streaming service that contained token-gated content available only to tokenholders (as well as people renting from said tokenholders). Ethflix implements an issuance schedule [e.g., 100,000 tokens initially with 1,000 tokens every month thereafter up to a max of 500,000 tokens, all of which are initially allocated to Ethflix’s DAO treasury — this is just one potential issuance schedule] for the tokens that can be used to access the token-gated content library and tells potential purchasers of the tokens that they will continue to produce and upload content for the exclusive viewing of those holding Ethflix tokens. Let’s say that people believe the team running Ethflix DAO because they have storied careers, impeccable reputations, and clear industry connections, etc. and that investors expect viewers, on average, would effectively be willing to pay around $10/month to access this content library — would the access tokens, or NFTs, not then be amenable to intrinsic, cash flows-based, valuation methods? Could not a fundamentally-minded investor build a model around the valuation of these hypothetical Ethflix tokens based on her expectations of the discount rate, expected life of the DAO’s content library, inflation schedule, and expected pricing power on the rental of the access tokens?

The combination of crypto-enabled value streaming, experimental auction schemes for token lending, and conditional, wallet-based discounts [lending platform could give rebate/discount conditional upon the borrower’s wallet (i.e., if a wallet has held someone’s social tokens or pieces of their NFT collection for >1 year then the content creator might want to give them preferential terms)] enables a level of granularity in price discrimination that’s difficult, if not impossible, to replicate by web2 internet companies, and certainly in a privacy-preserving way [Metamask doesn’t require your name and address to transfer tokens on your behalf whereas Netflix and Visa do; zero-knowledge proofs can allow for things like wallet-based discounts while credibly preserving end user privacy]. Whether extremely granular price discrimination that maintains user privacy represents “innovation” or “disruption” or “the future of media” or whatever is besides the point — the point is that a platform that can do effective price discrimination without introducing unnecessary user frictions and clunky monolithic implementations possess structural advantages against platforms that can’t, all other things being equal. Whether or not this represents a desirable future with respect to internet services [that ”price discrimination” based on your “wallet’s contents” might sound off-putting is understandable, though you could argue that’s basically what marginal tax rates are] for the average internet user is an open question and not the one I’m focused on here. In this space, everything that can be done (and is potentially profitable)

I’m arguing that an NFT in the context I’ve just roughly outlined would undeniably have an intrinsic value in much the same way that equity of Netflix has intrinsic value. I’m aware that the context that this hypothetical “Ethflix access NFT” inhabits is divorced from the currently existing NFT space that emphasizes a culture of hype, spectacle and speculation on image collections and the emergent cryptocultures around them — this software/media-subscription valuation framework does not apply latter class of NFTs. Once again, I am simply arguing that there will exist NFTs that clearly possess intrinsic value that is derived from the promise of future cash flows that the access they provide can produce.

There’s another, independent claim [i.e., Whether or not you agree with this claim should not affect your view of the previous claim] I’m going to make, which is that crypto-related mechanisms like tokenization and token-gating affords for superior digital experiences that are difficult, or even impossible, for traditional web2 companies to provide in steady state [i.e., assuming a world in which web3 media reaches comparable scale to web2 companies that don’t pivot or adapt, which is admittedly unrealistic]. Let me first talk concretely about implementations or potential implementations along the vein of “Ethflix” and hopefully go part of the way in supporting this second claim of mine.

Lit Protocol allows for conditional permissioning and access control based on on-chain data (e.g., holding a particular NFT, owning a token that indicates membership in a DAO, holding at least 1 ETH for 1 month, etc.) for media files (images, videos, pdfs, etc.), raw data, dynamic content loaded from a server (e.g., a livestream), a Zoom meeting, a Gather room, a Google doc, etc. Livepeer is a decentralized video streaming network that incentivizes miners to run Livepeer nodes within a decentralized network of GPUs used for transcoding video. reNFT is protocol that enables p2p (p2p = peer-to-peer, as opposed to peer-to-protocol) lending and borrowing of NFTs which allows renters to enjoy usus rights of the token (e.g., access to a token-gated experience) for the borrowing duration while preserving fructus and abusus rights of the token lender. I’ve previously seen demos of Lit Protocol being used in conjunction with Livepeer for token-gated livestreams [I don’t feel like searching for them right now, so I guess you’ll just have to trust me for now] so it’s not a stretch to imagine a business model in which livestreamers can token-gate exclusive streams and/or interactive experiences that require their social tokens to access, thereby creating demand for rental of said tokens and imputing an asset value on the tokens.

But there’s room for experimentation here on all levels. The streaming layer could be hosted by Theta Network instead of Livepeer, or it could simply be hosted on a more centralized platform like Cloudflare [I’m unsure to the extent of Cloudflare’s integration with web3 services, but given that they’ve announced that they will serve as a validator node for Ethereum and that they’re looking to claw away incremental compute/network/storage demand from the Big Three hyperscalers, it’s very likely that NET services will be crypto-friendly]. Additionally, there are already a myriad of different NFT lending platforms beyond reNFT — you can bet that they will eventually seek to differentiate themselves as this market grows by doing things like integrating different auction mechanisms that can improve price discovery and liquidity for these access tokens.

Or it can be an entirely different architecture that uses a protocol like Superfluid, which allows for programmable streams of value (e.g., tokens, stablecoins, etc.) between wallets that are connected by a “Constant Flow Agreement.” DeFlix and StreamRent were two Superfluid hackathon projects that demonstrate some of the more exotic designs that are possible within the crypto ecosystem — the former allows for pay-per-second video streaming by making videos into NFTs and the latter enabled automated repossession of NFTs at the termination of a stream [actually, I haven’t diligenced whether these projects worked or not because I can’t find them online, but I’ll just take their word on it].

, will be done.

So let’s put it all together together now. Let’s say I have 1 of 1,000 “Nvidia Earth-2” NFTs that are required to access the token-gated Nvidia supercomputer running Earth-based physics simulations in the cloud. And let’s say that spending time in this simulation environment is valuable for all types of people such that I can expect to sell access (via renting my access token) at an average rate of $10/minute which, assuming 100% utilization and that the average $10/minute rate holds, annualizes to $5,256,000/year. The fundamental investor doing asset valuation would recognize that this NFT is intrinsically worth at least $5,256,000 based solely on this year’s expected [minute-to-minute] cash flows. The conversation quickly becomes “How many years will these supply-demand dynamics hold?”, “How long will it take for substitutes for high fidelity simulation environments come into the market, either through Nvidia putting more supercomputers online or through lower fidelity environments that maintain most of the utility?”, “How will the minute-by-minute pricing power of these passes change? Will the demand from the researcher cohort using the simulation environment eventually fall off after they’ve ran the majority of their experiments?”, etc. — these are questions about cash flow growth [let’s assume Nvidia embedded their expected costs, compute and otherwise, when initially selling those 1,000 tokens] and what multiple an investor should slap onto these cash flows based on expectations of useful life and pricing power. Let’s say that you work out that a tokenholder can rent their token at $5,000,000/year for 10 years (i.e., assuming no growth in cash flows), residual value is $10,000,000 (i.e., An abundant, perfect substitute comes around in 10 years exactly, but the NFT itself can expected to still be worth 10mm as a flex), and a discount rate of 10%. I’m not going to do the math but the point is that this would be an appropriate framework for intrinsic valuation if this hypothetical token existed. It would constitute the intrinsic valuation of an NFT.

My example is convoluted but not entirely so — it would make sense that Nvidia provided limited access to use of their Earth-2 supercomputer because, unlike the cloud and edge computers used by Netflix to operate their streaming network, the hardware to run Earth-scale physics simulations is extremely scarce. And it’s clear that access to this simulation environment has undeniable pricing power for obvious reasons (e.g., Company A wants to see impact of climate change on their commercial real estate, Government B wants to scenario test potential floods and earthquakes, Researcher C wants to research whatever, etc.).

So there’s a rental market for the actual utility provided by the token but there’s also a market for ownership of the NFTs themselves in which these tokens are treated as capital assets that have cash flow producing capability. You could imagine that there could also be a futures market to lock-in usage rights for certain times of certain days. Or that some people would be willing to pay a premium for the Right of First Refusal on particularly high demand times. Or that some token owners agree to allow buyers to take on something like a capital lease for their tokens, granting the leasee the right to buyout the token after some usage criteria has been met. Maybe you could tranche up the token by wrapping it in a contract that divides up ownership rights into an equity-like portion and a debt-like portion (not to mention hybrids or share classes). And so on, and so on. You get the point.

I’m not going to argue whether or not this hyperfinancialization of increasingly obscure, dematerialized assets broadly represents “progress” but I can argue there exist designs of digital experiences that can clearly make use of the mechanisms I’ve mentioned thus far. You could imagine some interactive digital experience, like a Minecraft server with voicechat, where participants have both read and write access in the environment. The server instance can only hold so many concurrent players on at one time, so it makes sense to limit access through something like a set of access tokens that can be rented out. Maybe the theme of this interactive experience is to optimize for the plurality of the server at any given time. Without diving into the weeds of what “plurality” really means, let’s just say that the ethos of this persistent interactive experience is to try to optimize for the number of people from different nations at any given time [assuming, of course, there’s some (ideally privacy-preserving) way for players to attest to their nationality, perhaps with a token linked to an online government registry] — we can imagine that players from a country that have 0 people online can programmatically be given a heavy discount for access and players from a country that is already overrepresented on the server have to pay a premium to gain access [what constitutes “over”-representation is up for the architect of the experience and the willing participants to decide for themselves]. Maybe this server requires some Proof-of-Humanity-esque attestation to deter people who would abuse write access (i.e., “griefing” in Minecraft) or harrass other participants in voicechat from thinking that they can get banned and simply create another wallet to access the server again.

From Decentralized Society: Finding Web3’s Soul:

Property rights are defined in the Roman legal tradition as bundles of rights to use (“usus”), consume or destroy (“abusus”), and profit (“fructus”). Rarely are all these rights jointly vested in the same owner. Apartment leases, for example, confer limited rights of use (“usus”) to the lessor, but not unfettered rights to destroy the apartment (“abusus”), sell it off (“fructus”), or even transfer use (subletting). Rights of real property (land) are typically encumbered by a range of restrictions on private use, grants of public rights of access, limits on rights of sale, and even rights of purchase by eminent domain. They are also typically encumbered with mortgages that transfer some financial value to lenders.

The future of property innovation is unlikely to build on wholly transferable private property so far imagined in web3. Rather innovation will hinge on the ability to decompose property rights to match features of existing property regimes, and code even richer elaborations.

Fees from token lending/renting for token-gated access are analogous to cash flows for usage-based software business models, thereby providing an intuitive valuation framework for tokens. The use of the term “token” rather than “NFT” was done intentionally for two reasons:

1. Token-gating mechanisms can accommodate fungible tokens, not just non-fungible tokens. In the same way that you can restrict access of a piece of content, Discord channel, livestream, etc. based on whether or not someone owns a particular NFT in their wallet, it’s also possible to restrict access to services based on whether this person owns 0.1% of the total supply of some ERC-20 token or whether this person has owned some quanta of an ERC-20 token for, let’s say, over a year.

2. The meaning of the term “NFT” varies wildly depending on who you ask to define it. Most people associate NFTs with pictures rather than the token standard that points to data that may or may not be an image — the medium is confused with the message. Furthermore, the emotional associations and connotations around “NFTs” often leads to critics and proponents talking past each other at most levels of discussion.

It is for these two reasons that, while I will primarily be referring to token-gating in the context of token-gating based on NFT ownership, I will use the term “token” instead of “NFTs” wherever I can.

But okay, here’s the point: token valuation methodologies in crypto, and especially around NFTs, are a vibrant area of contention due to the lack of cash flow streams (or, at least, the ability to “turn on” cash flows, which would be somewhat analogous to a SaaS company maturing through its company lifecycle deciding to de-emphasize growth and market capture in exchange for higher margins and FCF) that can be discounted back into the present for most cryptoassets. It’s argued that certain cryptoassets can be valued through the discounting of value flows back into the present because they are producing or will produce natural yields as a byproduct of staking (e.g., PoS Ethereum post-merge) or the protocols that the cryptoassets represent equity and governance ownership in either charges fees (or contains the ability to charge fees and alter take rates) for services rendered (this is essentially the argument for why governance tokens for DeFi protocols, like Uniswap, have value).

Although there exist NFT collections that enable holders to more NFTs through voting mechanisms or breeding (Nouns being an example of the former and Axie being an example of the latter), I don’t believe investors in these collections make valuation claims based on any DCF analogues. There’s perhaps an argument to be made that the BAYC collection and ApeCoin might represent intangible assets that maybe could be valued based off of some expectation of future merchandise, digital content, and event sales made by leveraging brand value [a business strategy exemplified by Disney’s leveraging of their characters and cinematic universes to drive content subscriptions and park tickets], but even this is an argument that represents a stretch for most traditionally-minded investors. For now, the most salient existing examples of NFTs that could be valued based on some expectation of future cash flows are those that correspond to virtual spaces (i.e., digital “land”) which can be rented by advertisers, thereby making them to analogous to traditional web2 advertising spaces (newsfeeds, articles, websites, search engine results, etc.) that can be valued using metrics like CPC, CPM, CPI, CPA, and CTR [some analysts have provisionally applied multiples-based real estate valuation frameworks around the similarity of the properties being assessed on digital “real estate” or digital “land” in (what I believe to be) an overly skeuomorphic manner].

All of the aforementioned valuation frameworks for NFTs I’ve just mentioned are speculative at best and would be readily dismissed (and probably ridiculed) by most investors who aren’t VCs given the prevailing attitude around all things NFT and Metaverse. I believe, however, that tokens that can be rented and used to access online services will undeniably have intrinsic value in a manner that clearly avoids the all-too-common self-reference that helps people justify speculation in the crypto space.

The most obvious example is that of a token-gated streaming service in which viewers subscribe through purchasing a token that corresponds to access to that day’s/week’s/month’s/year’s content — there are undeniable parallels between valuing the associated tokens of this hypothetical service to the valuation of media companies like Netflix and Spotify that traditional investors are already comfortable with. A skeptic following the argument up to this point might object “Why would a media organization choose to jump through all these hoops of token-gating their content and then issuing tokens, I don’t see the point” or “A media company utilizing these mechanisms will be outcompeted by centralized media services due to difficulty of customer acquisition as well as economies of scale in content production” or something along these lines — I must emphasize that the competitiveness of using token-gating mechanisms is a separate consideration from the recognization that valuation methods from software and emdia map well onto this hypothetical token-gated streaming service. Put even more concretely, if Netflix were to switch their subscription log-in and billing/payments systems architecture exclusively to a token-gating and token issuance model [they wouldn’t and they won’t for inumerable reasons, I’m just saying IF they did], then the value of the tokens involved would be undeniable.

Let’s make our example less obvious and less familiar. Let’s say that Netflix minted a limited supply of 10,000,000 access tokens and could rent these access tokens out for potential viewers of their content library [yes, they wouldn’t do either of these things, but please bear with me]. And let’s say that viewers could rent by the second such that they effectively pay-to-view on a second to second basis depending on how long they rent out the token that grants them access to Netflix’s content library. Therefore, someone who only watches one show during one month is charged significantly less than someone who binges two seasons and a movie every week of the same month simply — this would be tantamount to usage-based pricing, with the usage in this case directly corresponding to time spent (and therefore indirectly linked to costs from cloud compute, content caching, maybe even content amortization). This hypothetical, token-friendly Netflix service could charge different token rental prices based on the content being consumed (i.e., rental of tokens that can unlock access to more recent movies and shows cost more (a concept already very familiar media streaming services and media analysts)), the resolution quality of the content, and/or the ownership of other tokens in the viewer’s wallet (if the viewer’s wallet contains a token issued by a school that indicates that she is currently a student or by a government that indicates she is a citizen domiciling within a certain country (again, a concept already very familiar to streaming services)), etc. In this case, too, it is undeniable that each of the 10,000,000 hypothetical tokens could be valued intrinsically based on some expectation of future cash flows by Netflix’s internal finance team — why they would choose to issue 10,000,000 tokens in hypothetical might stem from a desire to moderate the variable costs from viewer demand on the necessary cloud computing and CDN services required to run personalization algorithms and stream content. But Netflix, like most internet services running at hyperscale on CSP infrastructure, already have predictive models for load balancing and pre-caching content in expectation of viewer demand, so, for this reason and others, the example I’m providing here is a stretch.

So let’s contort this example one last time, to make it less familiar and more crypto-native. Let’s say there was a DAO called “Ethflix” (pronounced “Eth”, not “Eth”) that operated a video streaming service that contained token-gated content available only to tokenholders (as well as people renting from said tokenholders). Ethflix implements an issuance schedule [e.g., 100,000 tokens initially with 1,000 tokens every month thereafter up to a max of 500,000 tokens, all of which are initially allocated to Ethflix’s DAO treasury — this is just one potential issuance schedule] for the tokens that can be used to access the token-gated content library and tells potential purchasers of the tokens that they will continue to produce and upload content for the exclusive viewing of those holding Ethflix tokens. Let’s say that people believe the team running Ethflix DAO because they have storied careers, impeccable reputations, and clear industry connections, etc. and that investors expect viewers, on average, would effectively be willing to pay around $10/month to access this content library — would the access tokens, or NFTs, not then be amenable to intrinsic, cash flows-based, valuation methods? Could not a fundamentally-minded investor build a model around the valuation of these hypothetical Ethflix tokens based on her expectations of the discount rate, expected life of the DAO’s content library, inflation schedule, and expected pricing power on the rental of the access tokens?

The combination of crypto-enabled value streaming, experimental auction schemes for token lending, and conditional, wallet-based discounts [lending platform could give rebate/discount conditional upon the borrower’s wallet (i.e., if a wallet has held someone’s social tokens or pieces of their NFT collection for >1 year then the content creator might want to give them preferential terms)] enables a level of granularity in price discrimination that’s difficult, if not impossible, to replicate by web2 internet companies, and certainly in a privacy-preserving way [Metamask doesn’t require your name and address to transfer tokens on your behalf whereas Netflix and Visa do; zero-knowledge proofs can allow for things like wallet-based discounts while credibly preserving end user privacy]. Whether extremely granular price discrimination that maintains user privacy represents “innovation” or “disruption” or “the future of media” or whatever is besides the point — the point is that a platform that can do effective price discrimination without introducing unnecessary user frictions and clunky monolithic implementations possess structural advantages against platforms that can’t, all other things being equal. Whether or not this represents a desirable future with respect to internet services [that ”price discrimination” based on your “wallet’s contents” might sound off-putting is understandable, though you could argue that’s basically what marginal tax rates are] for the average internet user is an open question and not the one I’m focused on here. In this space, everything that can be done (and is potentially profitable)

I’m arguing that an NFT in the context I’ve just roughly outlined would undeniably have an intrinsic value in much the same way that equity of Netflix has intrinsic value. I’m aware that the context that this hypothetical “Ethflix access NFT” inhabits is divorced from the currently existing NFT space that emphasizes a culture of hype, spectacle and speculation on image collections and the emergent cryptocultures around them — this software/media-subscription valuation framework does not apply latter class of NFTs. Once again, I am simply arguing that there will exist NFTs that clearly possess intrinsic value that is derived from the promise of future cash flows that the access they provide can produce.

There’s another, independent claim [i.e., Whether or not you agree with this claim should not affect your view of the previous claim] I’m going to make, which is that crypto-related mechanisms like tokenization and token-gating affords for superior digital experiences that are difficult, or even impossible, for traditional web2 companies to provide in steady state [i.e., assuming a world in which web3 media reaches comparable scale to web2 companies that don’t pivot or adapt, which is admittedly unrealistic]. Let me first talk concretely about implementations or potential implementations along the vein of “Ethflix” and hopefully go part of the way in supporting this second claim of mine.

Lit Protocol allows for conditional permissioning and access control based on on-chain data (e.g., holding a particular NFT, owning a token that indicates membership in a DAO, holding at least 1 ETH for 1 month, etc.) for media files (images, videos, pdfs, etc.), raw data, dynamic content loaded from a server (e.g., a livestream), a Zoom meeting, a Gather room, a Google doc, etc. Livepeer is a decentralized video streaming network that incentivizes miners to run Livepeer nodes within a decentralized network of GPUs used for transcoding video. reNFT is protocol that enables p2p (p2p = peer-to-peer, as opposed to peer-to-protocol) lending and borrowing of NFTs which allows renters to enjoy usus rights of the token (e.g., access to a token-gated experience) for the borrowing duration while preserving fructus and abusus rights of the token lender. I’ve previously seen demos of Lit Protocol being used in conjunction with Livepeer for token-gated livestreams [I don’t feel like searching for them right now, so I guess you’ll just have to trust me for now] so it’s not a stretch to imagine a business model in which livestreamers can token-gate exclusive streams and/or interactive experiences that require their social tokens to access, thereby creating demand for rental of said tokens and imputing an asset value on the tokens.

But there’s room for experimentation here on all levels. The streaming layer could be hosted by Theta Network instead of Livepeer, or it could simply be hosted on a more centralized platform like Cloudflare [I’m unsure to the extent of Cloudflare’s integration with web3 services, but given that they’ve announced that they will serve as a validator node for Ethereum and that they’re looking to claw away incremental compute/network/storage demand from the Big Three hyperscalers, it’s very likely that NET services will be crypto-friendly]. Additionally, there are already a myriad of different NFT lending platforms beyond reNFT — you can bet that they will eventually seek to differentiate themselves as this market grows by doing things like integrating different auction mechanisms that can improve price discovery and liquidity for these access tokens.

Or it can be an entirely different architecture that uses a protocol like Superfluid, which allows for programmable streams of value (e.g., tokens, stablecoins, etc.) between wallets that are connected by a “Constant Flow Agreement.” DeFlix and StreamRent were two Superfluid hackathon projects that demonstrate some of the more exotic designs that are possible within the crypto ecosystem — the former allows for pay-per-second video streaming by making videos into NFTs and the latter enabled automated repossession of NFTs at the termination of a stream [actually, I haven’t diligenced whether these projects worked or not because I can’t find them online, but I’ll just take their word on it].

, will be done.

So let’s put it all together together now. Let’s say I have 1 of 1,000 “Nvidia Earth-2” NFTs that are required to access the token-gated Nvidia supercomputer running Earth-based physics simulations in the cloud. And let’s say that spending time in this simulation environment is valuable for all types of people such that I can expect to sell access (via renting my access token) at an average rate of $10/minute which, assuming 100% utilization and that the average $10/minute rate holds, annualizes to $5,256,000/year. The fundamental investor doing asset valuation would recognize that this NFT is intrinsically worth at least $5,256,000 based solely on this year’s expected [minute-to-minute] cash flows. The conversation quickly becomes “How many years will these supply-demand dynamics hold?”, “How long will it take for substitutes for high fidelity simulation environments come into the market, either through Nvidia putting more supercomputers online or through lower fidelity environments that maintain most of the utility?”, “How will the minute-by-minute pricing power of these passes change? Will the demand from the researcher cohort using the simulation environment eventually fall off after they’ve ran the majority of their experiments?”, etc. — these are questions about cash flow growth [let’s assume Nvidia embedded their expected costs, compute and otherwise, when initially selling those 1,000 tokens] and what multiple an investor should slap onto these cash flows based on expectations of useful life and pricing power. Let’s say that you work out that a tokenholder can rent their token at $5,000,000/year for 10 years (i.e., assuming no growth in cash flows), residual value is $10,000,000 (i.e., An abundant, perfect substitute comes around in 10 years exactly, but the NFT itself can expected to still be worth 10mm as a flex), and a discount rate of 10%. I’m not going to do the math but the point is that this would be an appropriate framework for intrinsic valuation if this hypothetical token existed. It would constitute the intrinsic valuation of an NFT.

My example is convoluted but not entirely so — it would make sense that Nvidia provided limited access to use of their Earth-2 supercomputer because, unlike the cloud and edge computers used by Netflix to operate their streaming network, the hardware to run Earth-scale physics simulations is extremely scarce. And it’s clear that access to this simulation environment has undeniable pricing power for obvious reasons (e.g., Company A wants to see impact of climate change on their commercial real estate, Government B wants to scenario test potential floods and earthquakes, Researcher C wants to research whatever, etc.).

So there’s a rental market for the actual utility provided by the token but there’s also a market for ownership of the NFTs themselves in which these tokens are treated as capital assets that have cash flow producing capability. You could imagine that there could also be a futures market to lock-in usage rights for certain times of certain days. Or that some people would be willing to pay a premium for the Right of First Refusal on particularly high demand times. Or that some token owners agree to allow buyers to take on something like a capital lease for their tokens, granting the leasee the right to buyout the token after some usage criteria has been met. Maybe you could tranche up the token by wrapping it in a contract that divides up ownership rights into an equity-like portion and a debt-like portion (not to mention hybrids or share classes). And so on, and so on. You get the point.

I’m not going to argue whether or not this hyperfinancialization of increasingly obscure, dematerialized assets broadly represents “progress” but I can argue there exist designs of digital experiences that can clearly make use of the mechanisms I’ve mentioned thus far. You could imagine some interactive digital experience, like a Minecraft server with voicechat, where participants have both read and write access in the environment. The server instance can only hold so many concurrent players on at one time, so it makes sense to limit access through something like a set of access tokens that can be rented out. Maybe the theme of this interactive experience is to optimize for the plurality of the server at any given time. Without diving into the weeds of what “plurality” really means, let’s just say that the ethos of this persistent interactive experience is to try to optimize for the number of people from different nations at any given time [assuming, of course, there’s some (ideally privacy-preserving) way for players to attest to their nationality, perhaps with a token linked to an online government registry] — we can imagine that players from a country that have 0 people online can programmatically be given a heavy discount for access and players from a country that is already overrepresented on the server have to pay a premium to gain access [what constitutes “over”-representation is up for the architect of the experience and the willing participants to decide for themselves]. Maybe this server requires some Proof-of-Humanity-esque attestation to deter people who would abuse write access (i.e., “griefing” in Minecraft) or harrass other participants in voicechat from thinking that they can get banned and simply create another wallet to access the server again.