September 2021 Newsletter: Why bother with NFTs?
In this month's edition:
My First Virtual Exhibition - with NFTs!
Is Lab-Grown Meat Viable?
Stripe is eating Quickbooks
Sagrada Barcelona - My first virtual photography exhibition
Last week, after a trip to Barcelona, I pulled together a virtual exhibition of photos and short videos. You can check out the virtual experience for free here - it's pretty short. Let me know your favourite!
This virtual art gallery pulls together a few technology trends:
Virtual Reality - a major area of investment for Facebook. Virtual Reality holds a lot of promise because the experience is far more immersive than looking at a flat screen. My virtual exhibition is a bit clunky, but within ten years I think we'll be viewing these with virtual reality goggles or glasses. Maybe within twenty years we'll have some kind of eye implant for immersive experiences.
Non-fungible tokens (NFT). I have turned each digital image/video into a digital token that can be sold or auctioned (example of one of mine here). NFTs are a big area of hype and speculation, so let me give a little background on where I think there is value.
My take on NFTs
Art is an alignment among a group of people, however big or small, that something is valuable for aesthetic reasons. Non-fungible tokens are unique digital identifiers for a given piece of art. For example, a digital photograph might have an identifying code of C789AB89D.
The list of different pieces of art along with their owners is called a blockchain. That blockchain/list is stored on the internet and anyone can download the list to see who owns what. If you own a piece of art that is listed, you can transfer ownership to someone else by signing off on a transfer with a passcode (called a private key).
So, NFTs are just digital codes for pieces of art, and a blockchain is just a public list of who owns what. Why is it useful to have a public list of art?
The list* is public and not controlled by any one government or company. The list is accessible to all people in all countries (if they have internet). Nobody can be blocked from reading the list, or contributing by receiving or sending tokens.
Typically when an artist sells a painting, they only make money up front - i.e. when the painting is re-sold, it's the owner of the painting, not the artist, that benefits. With digital ownership lists, artists can get a percentage cut of the transaction every time their art is sold.
Can other people copy my digital image for free?
Using digital lists/blockchains and NFTs is not a replacement for copyright laws. Yes, somebody can use a digital image that you own, but they would typically be breaking copyright and usage laws in many countries. The list/blockchain doesn't enforce any of these rights (although they allow you to check if a given image is owned by someone else).
For example, if I buy an NFT for a Mickey Mouse picture, typically I would own the rights to that image (and the list/blockchain can prove that), but for any enforcement of my rights, I would to resort to state laws.
Where is the digital photo associated with my NFT stored?
The art/content/photo/video itself is not stored on the blockchain. The blockchain is just a list of items and owners. The image itself is typically stored by the marketplace where you buy it (e.g. opensea.io). Further, it is common for the same photo to be stored in many places/marketplaces across the internet, and owning the NFT gives you ownership of all of these copies. Here's how this works:
Lets say you own a photo called BigLabrador.jpeg , which is a digital file.
A digital identifier (i.e. the NFT) is created for your photo by taking the digital file and doing a calculation (called a hash) on that file's data to come up with a unique code - let's say BA6789EA. This code is your NFT.
There may be many identical copies of that NFT on the internet. If they are identical, then doing the hash calculation on the digital file will lead to the same code BA6789EA.
By owning an NFT, you own all of the files that have a certain unique code.
In summary, public lists for ownership of things makes good sense because a) they provide the best access to everyone and b) it's easier to work with digital records that are global than manual or siloed records that are just controlled and/or formatted by one government or country. With public ownership lists, you can have global lists of art, of houses, of cars etc. that are accessible to everyone.
* There are actually a number of lists/blockchains you can choose from if you want to put your art on a public list. The most common right now is called Ethereum. It is possible that different protocols will specialise in different types of list (e.g. art versus property). It is also likely that there will be one protocol that will dominate (currently Ethereum is dominating for NFTs).
Is lab grown meat viable?
I like the taste and nutrition that meat provides. I'm an optimist on technology and feel we have a chance of making lab-grown meat that tastes as good as meat from animals.
This article on the cost of lab grown meat is highly skeptical that it can be anything more than a very high end product, and forced me to reconsider my optimism.
The core argument of the piece is that lab grown meat involves scaling up a pharmaceutical grade process orders of magnitude larger than has been done before and - even under aggressive assumptions - any reasonable estimate at artificial meat costs comes in far above the price of meat we eat today.
One framework for thinking about the cost of lab grown meat is to compare it with the efficiency of animals we breed. Knowing it takes about 10 calories of feed for every calorie of chicken protein or 25 calories of feed for every calorie of beef protein, we can compare these numbers to the calories of energy it takes to make one calorie of artifical protein.
A 2021 study by CE Delft, they considered a baseline scenario for a large scale facility producing 10,000 tons of artificial meat per year. They estimate a need for 220 million kWh of electricity to run this plant each year. This works out to about 22 kWh per kilogram of meat (which is about 19,000 calories). One kg of meat (let's say beef) has about 2,500 calories. So, the study's baseline scenario for scaled up artificial meat production is about 7.5 calories of electricity for every calorie of artificial meat delivered. This doesn't include energy for raw materials or transport, but doesn't seem entirely awful compared to breeding chickens or cows.
Unfortunately, the baseline price estimate for at-scale artificial meat is well above $10,000 per kg of meat when all costs are included. This begs the question of how companies like Impossible Foods are surviving.
Impossible Foods is a private company, so there is no data available.
Articles covering the environmental efficiency of artificial meat commonly do not provide references to raw data making it hard to verify energy and costs.
My assumption here is that the capital cost of building factories to make the artificial meat - and possibly the energy and emissions in building these factories - are not fully being built into the footprint and cost of the meat. When impossible burgers is selling burgers at $12 in a sandwich shop, but a 2021 study puts the costs at over $20k per kilogram, there must be something hidden somewhere.
In conclusion, I'm somewhat optimistic we'll crack artificial meat over a long period of time - maybe 50 years, but in the meantime I suspect early companies may become vegan guinea pig casualties of innovation.
Stripe is Eating Quickbooks
In the US, it is common for startups to use Quickbooks to manage their accounts. Quickbooks is software that connects to bank accounts and allows you to tag spending and revenue in order to generate financial reports. In Ireland it seems that startups tend to use Xero - which is similar. Quickbooks is a big business that is owned by Intuit, which also owns TurboTax - the most popular tax software in the US - and also Mint (a money management tool).
Stripe is a company that allows you to accept credit card payments. It was founded by two Irish lads, John and Patrick Collison. Stripe is offering more and more products to businesses. For example, you can set up a business using Stripe Atlas. You can automate calculation of VAT (or sales taxes) using Stripe. Just a week or so ago Stripe announced that you can now generate accounting reports using Stripe Revenue Recognition. Stripe is including everything Quickbooks does, but their software is easier to use and they are quicker to launch new products. Most of Stripe's customers started as startups, but those startups have since gotten big, so Stripe is growing with their customers and they are eating up Quickbooks.
All of Intuit (including Turbotax and Mint) is worth $150B today. Stripe is worth just under $100B. So, in some ways Stripe is already worth more than Quickbooks.
It is interesting that Stripe is not supporting cryptocurrencies. They did for a while but then stopped. I feel two ways about this:
The Stripe founders are very smart. If they aren't building solutions for accounting on blockchains/lists/crypto, maybe blockchains/crypto is less good that I think it could be.
Maybe the biggest risk about Stripe is that they are wrong about crypto.
That's it for this month! If you'd like, I can send you this newsletter each month. Just sign up with your email below: