Decentralised orgs 2/x - tokens
Next: DAOs
In the previous post we introduced the key concepts of blockchains and had some examples of services that it makes possible. We ended up with an examples of decentralised autonomous organisations.
TL;DR Decentralized platform use tokens in a multitude of roles: as a way of counting contribution value, as payment assets, to record ownership or access rights and so on. Tokens (or actually the smart contracts controlling them) are programmable, meaning that the rules can be whatever designers desire.
Tokens
It’s time to peek at tokens that are the programmable mechanism of keeping track of state of the system. The state could be for example value that you use to pay for products and services, decision power used for votes, rights to access e.g. a facility or hotel room. The programmable nature means that tokens can have all kinds of constraints or additional features not commonly available in the centralised environment. There can be location, time, holder or recipient related constraints or non-linear rules for how for example votes are calculated. Anything that can be programmed, is possible. Whether such changes make sense or will users be attracted to such schemes is another thing, but the design space is much larger.
For different uses, different token types.
The most commonly known are native blockchain tokens like Bitcoin and Ether that are used to pay for the use of the blockchain basic functionality – i.e., making transactions. You can use them naturally to pay for other services as well but as free floating entities they are not good as payment vehicle. There is no institution supporting their value like central banks for national currencies.
For every transaction on chains like Bitcoin and Ethereum the caller adds a small fee that the miner collects in addition to the reward they get for creating the next block. Originally introduced in Bitcoin because the supply of Bitcoins is limited to 21 million and at some point the miner rewards stop being issued and the only reward will then be the transaction fees. Other chains have adopted this convention to prevent denial of service attacks where somene would flood the network with transactions.
Other chains do not have necessarily have a limited supply. In Ethereum the case is even such that after the transition to poof of stake the supply may start reducing. technically the transaction fee will be split into tip for miner and base bee and the base fee tokens will be destroyed aka “burned”). According to simulations the amount of tokens will start decreasing about 2% yearly.
Many blockchains also have mechanisms to implement your own tokens on top of the basic chains. For example, on the common Ethereum chain, you can use smart contracts to implement your own. In Ethereum there are de facto standards for implementers on how they need to look like. ERC-20 for usage/governance and ERC-721 for collectibles aka NFTs, more on those later. And to make it easy, open-sourced implementations exist. More on those token types later. These types of tokens that are implemented on top of a blockchain are called L2-cryptocurrencies or L2-tokens for short.
A cryptocurrency can be pegged (tied) to a national currency. Most commonly to US dollar. Pegging means that you buy one token with one dollar and sell for one dollar with small transaction fees for the service. Simplest implementation is having corresponding amount of national currency (usually USD) in a bank account. These are called stable coins as their rate of exchange is fixed. They make for a great option for handling payments.
Some projects have tried to create algorithmic tokens that would be based on a mix of assets, typically other crypto tokens of a kind. These have a feature that when the whole market goes down, they cannot hold their peg and hence have experienced loss of trust from customers.
Stable coins are an example of asset backed tokens (also called commodity tokens). As a general concept they are pegged to some standard asset class that have value like gold, oil or national currency. There are regular audits by a trusted third party validating that the assets are really purchased and in safe custody.
Utility tokens are used to consume specific service – say to purchase computing, storage, data transfer or water as examples. They are identical to lunch coupons today or pollets that people used to pay for amusement park rides as example. Each service has their own tokens that you usually buy with other cryptocurrencies or with credit card or bank transfer.
Easiest to buy tokens is via centralised token exchanges. These support purchases using national currency. To be able to purchase tokens you need to go through identification process (KYC - know your customer) where they verify your identity. This functionality of going back and forth between tokens and national currencies is called on-ramp/off-ramp.
Decentralized exchanges allow you to buy tokens with other token types.
As an example of a decentralized web3 service using utility tokens, Golem is selling computing capability. Their token is called Golem Network Token or GNT and people who have computing needs like image rendering for movie effects, protein folding or any general computing use GNT to pay for submitted jobs. And people who share their computing resources gain same tokens as rewards.
Governance tokens give voting rights to guide the future of a given project. An increasingly common scheme today is to distribute governance tokens to stakeholders of a project. This means that people who provide the assets can get both utility tokens as payment for their contribution and also voting rights as governance tokens.
We’ll return to the topic of different models how governance tokens can work later.
Purpose driven tokens are a newer class of tokens. They are used by project or sometimes by cities to incentivise behaviour the project wants to promote. This can be activities like using public transport, or some volunteer activities etc.
The latest flavour of the month entering the token family are so called NFTs. To explain, a little backgrounder first.
To understand token characteristics a little bit better, we can split token into two groups: ones where all tokens are alike (one Bitcoin is just any other) or not (my certification for electric work allows only me to do this work). Or tokens that can be split into smaller units (you can own a fraction of a Bitcoin but owning a fraction of a certification does not make sense)
Crypto-communities use terms like fungible and non-fungible for these two types. (For a non-English speaking person these sound like a disease that infects the skin between your toes and start spreading with alarming speed).
Different types of tokens can be presented in the following grid.
Non-fungible tokens (NFTS) are unique. For example, a university degree would be non-fungible. Similarly, many predict that shares in a company and property will be electronically recorded in future and these could well be implemented as NFTs. Some NFTs are non-divisible and others not. For example titles and degrees cannot be divided but a lot of land can have multiple owners with different fractional ownerships.
Equity tokens represent shares in a company giving you a share of the future profits and decision rights in a bundle. When equity tokens are offered to the general public, they fall under the normal financial regulatory laws. Organizations offering equity tokens need to register their security in every county where people would be allowed to buy it and follow local legislation that tends to differ from region to region.
The legal complications mentioned above in equity tokens mean that in practice this token type is not common. Often it is not clear if something would fall under the regulatory eye or not so all crypto projects need to keep watching the regulatory space.
One particular type of NFTs have become popular recently – namely ownership to digital art or digital memes. Some of them have been sold by their original creator at very large prices in NFT auctions. Technically they just contain a link (URI) to meta data that links to the original picture. This sold sometimes for millions. Art markets follow their own logic. One can also view them as membership fees to a club because some NTF collections enable owners to members-only discussion forums or events just for members. If the community is active and has prestigious members, the NFT aka membership entry fee can rise high. Whether these hold value over time is anyone’s guess but value they are given by the market for now as such is the fleeting ethos of our time.
In general, tokens allow unbundling rights that used to be together in the past and programming new logic around them.
Decision power and access to future profits of a company can be separated. A company can tokenize just one large projects and sell up future profits from it thus gaining revenue faster than what the payment posts of the project say.
Dividends could be issued monthly or even weekly if it made sense. Different products or services could have their own crypto-tokens.
An artist might reserve 15% of the value increase in each subsequent sale to go to them. This naturally works only if sales happen via smart contracts but as they maintain the providence at the same time, there is incentive to do so.
Example: IP-NFT
Now that dismissing NFTs is in season, it’s time for two NFT examples, just for good measure. Specifically, IP-NFTs. IP standard for intellectual property and refers to patents, artworks, designs, symbols, logos, names, images used in commerce etc.
How would NFTs work here?
NFTs are just tokens residing in a blockchain that can be transferred between addresses. Optionally they may have a URI (link) to meta-data that resides somewhere else. For images there is specification for the meta-data file structure but different NFTs classes may have their own file structures that suit them better. The standard for NFTs is quite readable for people with some programming exposure (your mileage may differ) and is here: https://eips.ethereum.org/EIPS/eip-721
IP-NFTs have their own meta-data and are used to make intellectual property transactable easily. They have an associated legal contract for using the IP that the buyer agrees to when they acquire it. The user may buy full rights or a license to the assets.
Say the buyer wants a subcontractor develop something for them or a university to conduct some test or research, the contract would say that the rights and the measured data sets belong to the buyer.
Alternatively, there can already be some assets available. These could be an early phase innovation or research idea, large data set of measurements from some study, discovery from previous work that needs further investment into full product, designs for products or medicinal molecule and the production instructions as examples.
With IP-NFTs these types of assets can be made digitally available so that the seller sets the rules and prices and transactions happen digitally. The parties naturally need to have go through so called KYC (know your customer) verification so that the legal parties are known.
Physically the data resides for example on decentralised data storage like IPFS or it could also be on a centralised server. This allows trading all the assets as one bundle (all documents, data sets, code, machine learning models, whatevers) in digital manner. The data is not clear text but encrypted and trading means handing over the keys to unlock it.
In simple terms an IP-NFT consists of a legal contract, smart contract for managing purchases, storage for the actual asset and meta data about the contents.
What makes IP-NFTs attractive is that they are part of a composable web3 ecosystem. The seller can use already existing decentralised finance services (later more in future posts on these) for example to set them up for an auction (e.g. Gnosis auction) or sell one-to-one for exclusive access or allow multiple buyers license to it.
In addition, web3 allows to fractionalize the IP-NFT. i.e., divide it into many parts and allow people to own only a small portion of it. One IP-NFT could be made into a say one million tokens to be sold to investors. The rewards will then be distributed to the owners in relation to their ownership. These tokens can naturally be used as collateral for raising loans or just easily sold up when funds are needed.
The token holders then can vote what to do with the IP.
And the emerging decentralised finance also allows people to create sell or buy futures over IP-NFTs. So, people who do not believe in the value of the IP can bet against it and gain reward if they are right and people who strongly believe can bet for value increase. The price of these futures gives some kind of market sentiment on the value of the IP. Whenever some positive data arises from ongoing work. the price for tokens will raise and vice versa for bad news.
A quite similar method of getting sentiment will be prediction markets where people stake value against or for something succeeding or failing.
Since the digital data is very easy to copy, what guarantees the seller can have that the buyer follows the rules set in the contract ? Technically there is very little to do (true also of the traditional trading of digital assets). It is naturally best to trade with known parties with reputation to lose if found guilty of unmoral activities.
Web3 allows some new mechanism for trust building. All of them are based the concept of having skin-in-the-game. If you do the right thing, you’ll get rewarded but if you misbehave, you get negative outcome. Typically, some amount of value (token) is used as the “skin” part.
Token curated registries is one option. In it participants maintain a curated list of entries with good reputation. They work so that buyers wanting to enter the list make a deposit together with their request to join the list. Token holders who also act as curators review it. If all are fine with including this party, they are added to the list and get their deposit back
Anyone can contest the entry of the applicant with rationale. This leads to a vote among participants. If applicant is rejected, they lose the deposit sum which is divided among the challenger and voters.
Token holders have an interest to make sound judgement because this will increase the price of their tokens. For example, just rejecting all new entrants would gain short term reward but would collapse the value of the list as an information source.
More here: https://medium.com/@ilovebagels/token-curated-registries-1-0-61a232f8dac7
Token curated registries are still new and will evolve in future. The above example does not as an example take the passage of time into consideration. A company that is now reliable is a different one in five years’ time after a series of slicing and dicing in mergers or acquisitions have resulted the company name landing on top of some random leftover.
So how are IP-NFTs better than the current cure?
Today development happens following a common funnel model starting with lots of ideas that go through a set of steps where on each step the number of concepts gets smaller (validation, project formation, development, launch). All of this is done inside organisations not sharing results or data with anyone on the outside. This means a researcher at a university with an idea needs to first start patenting it and raising money for a startup and trying to understand the university’s rules for spin-offs while not even knowing if the idea works or has market need. Inside an enterprise the process is similar with the need to start building a business case for something that does not exists yet (meaning no analyst info exists and no sales team is foolish enough to commit any sales figures). Alternative is going for own startup. This is a slow and expensive way of innovating. But truth to be told, often existing managers prefer it as their primary target is to protect the status quo and innovations just drain their budget.
If the idea turns out not to work, there may still be substantial amount of money left in the startup and people wanting to keep their jobs will find new use for it. This is called “pivoting” in the startup world.
IP-NFTs allow financing new ideas at a very early stage with much smaller sums without the idea creator first patenting nor raising startup level financing from venture capitalists. The idea itself is stored in immutable format on blockchains proving who had prior art in case of disputes. Transacting on the IP alone for validating or disproving a thesis is much smaller task than hiring a full development team for the expected full productisation project.
This happens at the same time when all kinds of lab equipment, development and research work and tools become available as a service via various cloud providers. Industrial design can be purchased easily and heavy investments into specialised manufacturing or research equipment can be skipped.
Second use case for IP-NFTs is around a concept called pay for success. Today’s monetary rewards give incentives for subset of all useful causes. For example, there are around 20k medicinal molecules off-license that can are produced and available for a very low cost. These medicines are known to be safe and doctors are familiar in prescribing them and on their side effects.
There is no incentive to look if these could be used for treating any other diseases. Even when there is some small-scale evidence from around the world where they have been successfully applied, no one wants to do the expensive final clinical trials to verify this because you cannot patent the results.
With IP-NFTs it is possible to set up monetary rewards for any team that successfully conducts tests on off-license medicine to find effective and dirt-cheap treatments for all kinds of diseases. Patient associations at least should be interested in this and may be able to raise the needed money from donations.
End-result is a totally meritocratic reward system allowing anyone globally to participate. Often money allocations done by expert panels strongly favor well-known research teams that closely follow the current consensus view of the domain. Pay-for-Successes only care if it works or not, ignoring who you are or what you are supposed to believe.
Pay for failure is a similar system where payment is tied to generation of negative data. Today when research is conducted and the results are not good, the results end up not being published. Only positive findings go into research reports or customer studies. At same time journals have shifted their focus from rigorous methodology to novelty in the pressures for keeping the fleeting interest of paying readers.
This has led to a large replication crisis in many scientific domains.
Payment for failure is an idea that certain factor (say 20%) of research funding in each domain is allocated for replication studies. If a replication project finds non-replicable results, publishes its study setup, data set and code used for analysis, then a challenge period starts. If successful (i.e., original result was not replicable), team gets a reward.
A variant is where funding if gathered based on a hypothesis that some research results or medical interventions are not as successful as the company claims as examples. But in principle against any strongly held belief of the market. Proving scientifically the hypothesis is valuable information because it will ultimately affect company stock price. It’s very much like insider-info, but independently done research crowdfunded by IP-NFT.
IP-NFTs for the win
One of the hypotheses of this blog series is as follows:
Almost everything can be represented as a digital blueprint
Like software did, any digital asset will become open source
The IP-NFTs can speed up the process as anyone can open a payment for success NFT for any product or service they want. I.e., first collect a lump sum for donors for a payment-for-success NFT to design say an automated vertical garden for growing vegetables or some kitchen appliance like coffee maker. Setter of the NFT can set up rules and state for example that the product needs to bee easy to manufacture locally, simple to repair, durable and recyclable.
Then NFT pays for a working design and finally makes it openly available.
Each such IP-NFT could set up 20% aside to an escrow of sort for someone who can prove that the design does not meet the conditions.
A move to a world where everything is a digital design and openly available would thus be possible via Payment-for-Success NFTs and the quality kept good via Payment-for-Failure NFTs.
As a summary these IP-NFTs are a way of organising work and doing stuff that matters with no underlying organisation or company behind. That is a fairly wild idea for anyone with years of grinding the millstone in a corporate setting.
Next: DAOs