Exploring Continuous Token Models: Towards a Million Networks of Value.

Simon de la Rouviere Blocked Unblock Follow Following Feb 9, 2017

I want to put forth new exploratory models for thinking about tokenizing protocols, decentralized applications & interest networks.

My gut feel is that some of the token designs follows a model taken from what we’ve been used to. It “looks” and sometimes “feels” like traditional ownership structures.

But perhaps we are thinking about it like people thought about the web initially -> “Hey, let’s put a newspaper on a screen.” The massive change only really came about when media used the new features of the systems (“Hey, why aren’t we publishing faster than a day? We don’t have to print things anymore.”).

I want to put forth more thinking for continuous token models.

The idea is that instead of pre-selling tokens during a launch phase, the tokens are minted as needed through various means. The tokens are then dispensed for services rendered in the network.

Why can’t we sell interest/access without a centralised intermediary in real-time?

The simplest continuous token model is:

Use ETH to mint a token according to a price set by a hard-coded algorithm.

The cost of the token is dependent on how many tokens are in circulation. Thus, over time, the cost of minting will increase if there’s more in circulation. The “entry ticket” gets more expensive, the more people that wants to get through the door. If there are more people currently through the door (and haven’t left yet), the cost will increase.

The tokens are dispensed (‘destroyed’) for actions/services in the network. This decreases supply, and thus the cost to mint a token. The usage of the tokens are not restricted to only dispensing it.

Tokens can be bought/sold in secondary markets if the cost to get in is too high OR the users don’t want to dispense it. Rationally, the trading will happen under the current “ceiling” price.

This part of the model deals with *how* the token is minted. The second part which will be discussed below is *who* gets the ETH for minting the token.

The variations of the model are all based around who the beneficiaries are of the ETH that is paid to mint the token.

The potential beneficiary models are currently: 1) No one: Burning ETH, 2) Independent (pre-paid) Dynamic Beneficiaries, 3) Voted Dynamic Beneficiaries, 4) A Multi-Sig Beneficiary, 5) A DAO Beneficiary.

No one: Burning ETH

This is the simplest model. In essence, one is exchanging an affordant token (ETH) into a specific-use token. The issue with this however is like trying to fund a company by instead of giving say $2 million to a start-up, you are burning that $2 million. Yes… By removing the $2 million from circulation, you *are* making everyone’s else money worth more (less supply, same “demand”), but that’s not how we fund companies. It seems inefficient. Ideally, the ETH should go to beneficiaries that foster the project and the usefulness of the newly minted token.

Independent (pre-paid) Dynamic Beneficiaries

This model works by allowing the person who mints the token to pay *anyone* they wish. Yes. Anyone. The “rational” interest of the the buyer would be to pay a beneficiary that will make their newly minted token worth something now and into the future. Thus one can imagine, say around, an Ethereum-based continuous token community, that one could decide to pay: 1) Vitalik, 2) The Foundation, 3) ConsenSys, 4) Truffle, 5) Maker, and so forth. These entities all make the token more useful.

The issue however with this model is avoiding the ability for people to pay themselves and keep minting a token. This isn’t necessarily bad as one could say that “I” believe, “I” make this token valuable. However, the more potential negative outcome is that one can squat/block a token: make it expensive to join, and then no one can benefit from the network. It’s like the equivalent to someone buying all the tickets to an event and not letting anyone else in and enjoy the event.

There’s many potential ways around this: from time-locks to burning ETH according to entities that provably foster the token. However, like Vitalik recently said: You need to consider the benefit of good-enough coordination, whilst accepting that it can potentially also allow bad coordination. Too much hamstringing and you are moving nowhere. Too much burning, for example, to mitigate bad coordination will also negatively affect good, honest coordination.

Voted Dynamic Beneficiaries

Instead of pre-paying a beneficiary, the ETH is sent to a contract that has a leaderboard. Using the minted tokens itself, the community “votes” (dispensing the token) on who should get the ETH. After a certain time period, the funds are distributed automatically to the leaderboard.

Multi-Sig Beneficiary

In this model, a group/entity receives the ETH and has to sign off who the funds go to. It can either be a centralized decision making system (“we believe these people should be paid for the token to remain useful”). Or you could combine it with the voted beneficiaries. The multi-sig entity then needs to decide if it will conform to the wishes of the community. If it is being obstreperous, the value of the token might drop as the multi-sig entity is not conforming to the wishes of its network.

A DAO Beneficiary

Similar to the multi-sig, but with dynamic voting. You have to somehow “sell” ownership in this DAO, essentially creating a limited supply voting token. Not entirely a fan of this model since it just recreates the same complexities and inefficiencies that the continuous token models seek to avoid.

What do I dispense tokens for?

It’s up the network/project/community to invent use cases for what these tokens can be dispensed for. An early, useful use-case is curation.

For example, let’s say Truffle (or any other open source project) creates a continuous token model. One can buy a Truffle token to dispense it on what features are important to develop and work on. The ETH is then (ideally) sent to Truffle developers when the tokens are minted. The dispensed token then is a skin-in-the-game signal to get the attention of the developers what should be worked.

It starts to attach to recent experiments on the value of attention markets. Watch Meher’s video on the attention web & check out Maciej’s work on Userfeeds.

A variation on the attention market/curation model put forth by Paul Kohlhaas, is using the tokens as a deposit for signalling. Instead of immediately dispensing them, it is only dispensed when the desired action is completed by the network.

Curation (as in the Truffle example) is useful, but eventually you can start dispensing these tokens for more specific services as well (specific to the project). For example, a project could use a continuous token model where the tokens are dispensed for external revenue generated by the project (effectively being a dividend scheme).

Value of the tokens?

The continuous token model still provides reasonable speculation and ownership interest. If tokens are minted but not immediately dispensed, then it’s an indication of interest that you want to use the token later. You can decide to buy the token now, at an early stage, for a discount to potentially dispense later when the network itself is bigger (and the relative value of the token has increased).

Meme Markets = Global Namespace + Permission-less Continuous Token Models

I got into thinking about continuous token models mainly to think about how one can mint interest in networks of value without permission and without the incumbrance of requiring a specific institution to do so. What would be possible if we can mint interest in *all* networks of value? All memes? It could allow intense, rapid, global economic coordination around common goals as long as people agree that the algorithmic design properly quantifies their interest.

In order, for say Truffle to make use of a continuous token model, they have to deploy the contract with the algorithms in it. However, if you add a global namespace, users can start using this system WITHOUT the permission of Truffle. Users would then be able to immediately start congregating around common focal (schelling) points in the same way people organise around hashtags currently (another variation on naming I had for these new kinds of coordination schemes are ‘hashtag organisations’).

If the community just starts congregating around #truffle, the Truffle team would start receiving ETH. They will receive more ETH if they conform to the wishes of their users, by looking at what they are dispensing tokens for in their community.

Now, imagine you have a namespace for coordinating around any common goals? #ethereum, #vitalikbuterin, #vladzamfir, #casper, #globalwarming, #pokemon, #ows?

You have the ability to quantify all information and their network effects (memes).

That’s the goal I’ve been aiming for. Mass tokenization of *all* networks of value. I’ve been developing this in my spare time at ConsenSys here -> https://github.com/simondlr/mememarkets. An MVP is coming soon on the Ethereum testnet to start testing assumptions. Tests are being written for the contracts and the front-end is being developed. The APIs are done.

If this works, we’ll see likely millions of new tokens minted, around a huge long tail of coordination of global interest groups.

There are a lot of assumptions to test. Time to figure out. At the very least, I think the direction warrants testing, and experimentation. It’s likely that *something* here will be useful!

Conclusion:

Continuous Token Models take a different approach to quantifying interest in networks of value (from protocols, to communities, to organisations, to people).

The result (hopefully) is:

Rewarding multiple beneficiaries for development of a project without their permission. Aiding social coordination around common goals. The value of the network is more naturally tied to interest (less magic numbers). Networks are allowed to grow and die naturally. Mass, global, economic coordination.

Thanks:

Jarrad Hope, Juan Blanco, Paul Kohlhaas, Niran Babalola, Jesse Walden, Meher Roy, Maciej Olpinski, Mihai Alisie, Nick Tomaino and Karl Floersch for feedback!