The creator attribution ratio
And why the current debate around NFTs is a false choice
Lately I find that people fall into one of two camps when talking about NFTs:
It’s a bubble of snake oil — these NFTs have no intrinsic value! I could just copy the content and give it away for free! What do I actually own when I spend thousands on this “token??”
The future is here! We now have a way to own and monetize the value of digital content. Buy now, or regret it forever!
The truth is neither. NFTs are just a stepping stone to better attribution of digital content. Many current NFTs are not well connected to the value of the thing they represent and will not retain their value long-term. The bubble will pop, to a degree. But the smartest experiments out there aren’t the ones focused primarily on making people rich, they’re the ones paving a clear path towards the future.
The attribution ratio
In “Come for the NFTs, stay for the royalTs” I wrote about NFTs as one step on a path of attributing the value of digital content to its owners and creators:
Ultimately this is about attribution; how well can you attribute the value that content generates back to its creators and owners. You can essentially put different industries on a scale from zero attribution to perfect attribution. Music during Napster was pretty bad, now it’s better. Digital art before NFTs was essentially zero, now it’s a bit better. But it’s still not great.
To state the obvious, attribution matters to creators of all kinds — scientists, writers, YouTube stars, etc — for the same reason. The better they can capture the value they created in the past, the more value they can create in the future. If I create a song and get money every time it is played anywhere in the world, I can use those funds to make more songs that get more plays. If I discover knowledge that leads to a new invention and I get a cut of the proceeds, I can use those funds to discover new knowledge. Weak attribution is why Napster and the piracy wars almost brought the music industry to its knees, and why the scientists who helped invent the Covid vaccines are still driving dilapidated Camrys while pharma execs fly private. Attribution matters.
To demonstrate weak attribution in NFTs in that previous article, I used the example of NBA Topshots, where the value of your NFT theoretically changes in proportion to the cultural value of the clip you buy, but only if a lot of other assumptions hold true as well — principally whether people still think NBA Topshots are cool “virtual trading cards”. The clip of Lebron’s dunk could still be wildly popular in the future even if NBA Topshots have totally fallen off and nobody cares about them. That’s weak attribution.
I’m not saying weak attribution is bad in all cases, in the Topshot example it’s really by design, in the same way that you don’t assume that when you buy a Lebron rookie card you somehow get a cut of his future salary. My point is that creators rely on attribution in order to build a feedback loop from the content they build to further investment and advancement of their career.
Given the importance of attribution for creators, you can spot another weakness in the attribution of current NFTs. Their value is typically based on the value of the content at a given moment rather than total value it delivers over time. To illustrate this, let’s look at a typical meme. And when I say “value” I mean the total economic value created by the content, but the same would apply to cultural/reputational value as well.
There was a time when basically the only thing I saw online was various — and hilarious — iterations of this photo of Bernie. That was almost certainly the peak value for that meme. You might be saying “there was no economic value of this meme - everyone used it for free!” You’d be right, but the thing is tons of economic value that wouldn’t have been built otherwise was created, and that’s what we’re measuring here. The billions of impressions this generated and were monetized via ads, the myriad shows and newspapers and companies that used it in their marketing somehow, etc etc. If you had perfect information (which we won’t ever have), you could even tie it to the economic value generated by the kids who were inspired by this to become prolific meme creators themselves.
This meme still floats around, but over time it will inevitably decrease. I doubt my kids will be super excited about this photo, though I’m sure those of us who were part of it will always have some affinity for it. Whatever that final resting value, that’s basically the value delta for this meme. The total sum of cultural value delivered by this meme is basically the area under the curve.
Now imagine the photographer of this photo immediately minted an NFT of it as soon as she snapped the shot, and that in this hypothetical world the value of that NFT follows exactly the cultural value. How are different actors incentivized?
The first thing to note is that the volatility of the graph leads to a real “buy low sell fast” vibe. That’s what leads to the “pump and dump” behavior in some areas of NFTs that turn people off. Put another way, nobody benefits financially from the total cultural value (the area under the curve). In contrast if you buy the licensing rights to the Red Hot Chili Peppers you get paid every time a song is played, even if they get less popular after you buy those rights. But if the Red Hot Chili Peppers discography was an NFT, you’d lose money in that same situation.
This is basically the “gotcha” argument of most skeptics. They believe that these graphs are even more spikey than the one above and basically will come back down to zero — no value delta and zero eventual net value of the NFTs. Just a lot of winners and losers along the way. They see the NFT landscape as a digital casino. And given the fact that memes are notoriously spikey, they have a point.
The thing is, that isn’t the ultimate vision for NFTs and new technologies for digital ownership. It’s just the stepping stone to being able to better attribute the value of public goods. So the “gotchas” are talking right past the zealots. It’s a false choice.
Memes and other public goods
The most common examples of public goods are things like public parks or air, and it’s probably not intuitive why internet memes often fall into the same category or why that matters. Let me explain.
First, why it matters. If you’ve been following along, you know my interest in crypto comes from my connection to science and what you could call the “public goods problem.” As I laid out in more detail in “Coase’s Penguin is learning to fly” the essential challenge stems from the fact that some public goods — namely public knowledge ala Wikipedia or science — have historically been very hard to invest in and prioritize relative to their value.
There’s much more detail in that essay, but to massively simplify the economic terms that define public goods, they are things where it’s hard to keep people from using them and they don’t get used up in the process.
The Bernie meme fits both criteria. Once it’s online, it’s basically impossible to keep people from copying it and using it as they want. And clearly, if I use the meme it doesn’t somehow exhaust the supply of that meme such that others can’t use it. It’s a public good.
So what gets people excited about NFTs isn’t some belief that you can fully capture the value that these memes deliver, it’s the fact that NFTs are generating millions of dollars of economic value around public goods. The photographer of this Bernie meme would have no hope of charging people to use it. Somebody would copy it and before long it would be everywhere. But she can mint an NFT and make some money.
If I asked you to put a price tag to the total cultural value created by this meme, what would you say? Millions? Hundreds of millions? And then what about the amount you think an NFT of this Bernie meme created by the photographer could get? Tens of thousands maybe? Whatever the answer, that would be the current attribution ratio. Still very low, but moving in the right direction!
I won’t get lost in the weeds of science but I care about attribution ratios because they also illustrate science’s biggest paradox. The attribution is low while the impact is — according to some economists far smarter than me — perfectly equal to total economic growth per person (~5 trillion per year). It’s that weak feedback loop that means scientists struggle to make enough money to live while essentially providing the backbone for all market innovation. If we were designing the perfect system for innovation, we’d make it so that being an amazing scientist would make you famous, rich, and happy. The public goods problem has mostly meant the opposite, and the same has been true of the creators and artists building digital content.
So again, the false choice is to view NFTs as an end state rather than a step towards better attribution. Just look carefully at some of the best current NFT experiments and you can see what I mean.
Better than it was, not as good as it could be
This week, Sari Azout launched an essay on Mirror which she described as the “first ever experiment in Attribution+ where we not only cite sources of inspiration, but route economic value to them.” She created an NFT for the article, sold it and split the revenue generated between all of the collaborators and those who inspired the essay (aka those she cited).
Even though if you really want to get technical her experiment might not be the first (Packy McCormick minted an NFT for an essay with the same element of splitting revenue between those who inspired the writing), she is absolutely correct that it is innovative.
Let’s imagine how perfect attribution for these essays would look. In that perfect world if somebody read Sari’s essay and was inspired to create a company, and that company made billions of dollars, Sari would get a cut. So would the people that inspired Sari’s writing. If the essay helped Sari land a new amazing job, then some portion of the earnings would also go back to the people she cited in the essay. Perfect attribution would also mean the readers who supported Sari to keep writing would share in the future wealth as well!
Clearly, perfect attribution will never exist. But better attribution is important, and these experiments are a step forward, if still leaving some room to travel:
The steps forward
Previously, collaborators and cited thinkers would have gotten very little benefit from any commercial upside of Sari’s essay. Now they share the wealth!
The essay can generate revenue for Sari and for these other contributors despite being fully open and with no advertising. Public goods funding ftw!
The room to travel:
In this and many other recent experiments, the NFT also comes with a bunch of other “stuff” like 1:1 conversations with Sari and access to other paid services. There’s nothing wrong with that, but it makes it a bit hard to know from an experiment perspective how much of the willingness to invest comes from the cultural value vs these exclusive perks. If she just auctioned off time with her 1:1, it wouldn’t be a very exciting experiment — that would just be a fairly standard transaction.
The economic value created by the NFT is not very clearly connected to the economic value created by the essay. Attribution is still somewhat weak. If someone is so inspired by Sari’s essay that they create a new paid newsletter just talking about NFTs that cites her work and makes millions, there’s no clear route for that wealth to flow back to this NFT.
It may seem like I’m splitting hairs here, but my point is that while it’s amazing that Sari can generate revenue for this public good in an entirely new way, the attribution ratio is still pretty low. The details of that are important because there are ways to continue moving towards better attribution even in this complex territory of public knowledge.
The debate is in the next steps
The crux of my argument is that NFTs and other new technologies are opening up opportunities for people to push towards a more profound vision — one where the total cultural value of digital content can be better attributed to those that own and create it. It’s not to say they think current NFTs will lose value — and this is where it unfortunately gets a bit meta — many believe that the current NFTs will have long-term value exactly because they will have been a stepping stone on an incredibly important path. They will be the rookie cards of the next iteration of the internet.
If you buy that, then the debate shouldn’t be whether you think Topshot could be better, it’s whether you think Topshot like experiments will be better. I do, and I think the routes to that place fall into a few main categories.
Bringing more economic value on-chain. Projects like Audius are essentially driving towards this, where ownership of digital content is directly tied to subscriptions to that content. The same will soon become much more prevalent in other media, like newsletters. So for example, instead of running this one-off experiment on Mirror, I’d bet Packy Mccormick will eventually sell ownership of his content empire via a social token or something similar and then as that empire generates revenue it will be returned to those shareholders, sort of like dividends.
Social tokens will have their moment in order to facilitate that movement. NFTs are great for auctioning off a single piece of content, but you need something more permanent to better attribute long-term value back to digital owners. Social tokens will continue to skyrocket in adoption to serve that need. So in Sari’s case, imagine ownership in her token dictates the cut of any future NFT she sells. If I believe in Sari’s future value, both her and I are incentivized to avoid the peaky incentive of NFTs (time the market, sell at the high point) and instead choose something that delivers better total value attribution. As Patrick Rivera told me in a recent conversation, “ERC20s and social tokens will be the money layer and NFTs increasingly will be the digital goods layer which provides status, identity, belonging, reputation, and access.” Also check out Brian Flynn for great thinking on this point and look out for SeedClub, the YC for social tokens, to be a total home run.
Ownership meets versioning. One of the truisms about public knowledge specifically is that the best projects are cultivated over a very long period of time. Brynn Hobart and Alex Danco have both referred to this as “homesteading” topics. Take Ben Thompson’s aggregation theory for example — he has published dozens of articles which essentially improve on the existing theory and refine it. It’s easier to imagine how versions of the same project could align owners to long-term value. It clearly wouldn’t be perfect (again, perfect doesn’t exist) but let’s say a single NFT is minted to represent the theory as a whole. To fund the creation of each new iteration, that NFT is sold and the profits are split between all of the previous owners. When I buy the NFT, I’m not betting on the value peak of the current iteration, I’m betting on how much value I think the theory will deliver in the future. Again it’s not perfect, but this would be a step in the right direction.
Those are just a few of the routes to better attribution of value that I’m aware of and excited about, but there are countless more. So getting lost in the current experiments isn’t the point. The point is to ask whether those experiments are the beginning of the end of the way things have always been.
Big thanks to Patrick Rivera and Rapha Menezes for the thoughts and edits of this piece. Both of those guys are almost frustratingly smart on these topics and will build amazing things – go check ‘em out.