In my last post I introduced some concepts for discussion. The main concept was the concept of using models as a facilitator of lateral thinking. Lateral thinking is in my opinion one of the main keys to the success of Bitcoin, of Microsoft, of Apple, of Google, of Amazon. For example Microsoft founder Bill Gates created (to my understanding) the business model for Microsoft which at the time was groundbreaking. What made Microsoft one of the most profitable companies in history was instead of going with the business model of video game consoles, or the business model Apple went with of proprietary closed hardware, they went with a highly focused approach of software first and then leveraged royalties with the agenda of putting both a PC in every home and a Microsoft Windows on every PC.
Now I will highlight also the Tauchain community does have some lateral thinkers. @Karov for example is known for using lateral thinking in his blog posts and he often cites ET3. ET3 has a very unique business model based around an open consortium of licensees. This business model puts the law on it's side and leverages intellectual property in a similar manner to how Microsoft used intellectual property but then when we think about how to do things in a decentralized context an open consortium of licenses becomes innovative.
Game Theory, Models, Trust
Game theory in my opinion is critical to the success of Agoras. In my opinion it's the single most important thing to get right for the marketing, for the business model, and for achieving trust. Many people like to use the word "trustless" when talking about crypto but in my opinion trust minimized is more accurate. Trust does exist though and in fact using blockchain tech we have the ability to create an environment of even more trust that traditional law and justice systems enforce. This is all possible by reputation economics where the rules of the network become socially enforced "laws" due to the fact that there is a cost to reputation or a benefit to reputation. To cooperate with the protocol is to benefit ones reputation and to violate the protocol is to harm ones reputation in accordance with the game theory.
Models are important too because if we imagine a particular aspect of Agoras with a vending machine model we can more easily discuss and think of the possibilities. For example a vending machine which outputs algorithms, software, answers to questions, decision support or recommendations, etc. We also for thinking about marketing can work with the slot machine model looking at how casinos work and various speculative traps, speculative stickiness, and more.
For example if we look at Dan Larimer's invention of pegged assets which became the basis for the Steem Dollar but also the basis for stuff like Dai, we would be able to see that by using speculation we can apply game theory toward all kinds of creative outlets. Pegged assets is only the tip of the iceberg as there is really no limit to what can be gamified, what can be the profit map for lack of a better phrase.
And because I cannot find a better phrase than profit map I'll go ahead and explain that one. A profit map as I'm using it here is simply a list of tasks or speculation actions an earner can take to achieve their goal (profit). So for example the profit map for the Steem Dollar would be you can either take one side of the trade or the other, you can sell Steem for Steem Dollars or you can sell Steem Dollars for Steem. Much more complex profit maps can be formed than this but in order to think this way we would have to think of the participants as agents in a multi agent system.
For example roles which include:
Earners. Earners are those who interact with the network to earn a living or a profit. Earners need ways to earn and the network can provide ways for earners to earn in a manner which benefits the network and supports the protocol. Speculators. Speculators you could possibly classify as a separate class or as earners, as they seek to earn or profit from speculation but they still can speculate in ways beneficial to the network, so these would essentially be gamblers. When we think about speculators, this class of earners will be the class for which the slot machine model makes more sense because speculation is gambling. The slot machine works on variable schedule of reinforcement for those aware of the psychological science behind how the model works. Alternatively the vending machine model works on fixed ratio(rate) or fixed interval (time) schedule of reinforcement which is to say they do task A and they are promised delivery of the rewards of fixed amount at time Y just like a vending machine. [Earners class can include miners, content producers, developers, traders/gamblers, etc]
Holders. This is a part of the network which I think is very important and vastly under appreciated. The holders are the most loyal and core demographic in most networks of this type. If we are talking about banks we would say the holders are the customers who put their money in the bank. Why should we expect loyalty from the holders if there is no interest paid to the holders? Bitcoin pays no interest to the holders and on top of that there is also no guarantee the Bitcoin price will go up, so holders buy, hodl, and hope. My current understanding is that based on stakeholder analysis and stakeholder management techniques if you want more holders you should reward holders for holding. This also has to do with the time value of money and opportunity cost that they give up to be loyal to the network by locking their tokens or holding long term.
Buyers. Buyers unlike earners are not getting the tokens from mining, or from speculating. Buyers are people who are asked to give up their Bitcoin or fiat in exchange for supporting the network or project or perhaps in most cases buyers want to become holders. This is why it's important to encourage buying with marketing, but also why it's so important to reward holders who bought (took risk) to become holders. Buyers and holders in general make up the core of most of these networks right now and without them there is no means for the network to continue. Another way to think of buyers is as the customer, and usually the whole product is geared toward satisfying the customer.
These are just three common roles in most crypto networks. Most networks have stakers (holders who are rewarded), or miners (earners) or developers (earners), or buyers (people who bought on Coinbase or Binance). Most of the time buyers = holders but if there is no reward for holding then buyer could become seller or speculator. To think about game theory the first step I think for Tau as a community would be to come up with a list of all known means of creating profit maps and or for better expression, the mechanisms the network can use to elicit certain actions from the participants.
For example you can use time locks like Steem. Steem innovated with the ability to lock your Steem for a fixed period of time. This creates a fixed schedule of rewards when the owner chooses to unlock their Steem Power. This innovation can be improved upon greatly. For example who says lockups have to pay out in a fixed interval or fixed ratio reward schedule when it can pay out in a variable schedule? In fact, you can set it up so that reputation influences these things such that if a person has developed a very bad reputation then some randomness is added so that maybe their fixed reward payments become variable or less regular (they still get the same reward but they would not know when). Of course there are many other things you can do with this too, such as reward pools which pay out only in variable ratio schedule (jackpot style).
Why not use Tau to collaborative construct and decide upon the most important innovations?
I call it the most important innovations, the game theory, the models, and the trust. Tau will allow for us to reach agreement on the models to use, the game theory, the reputation structure, the rewards and costs, these are all going to be critical. At the same time many in the Tau community mention the need for marketing and there is no reason marketing cannot be done decentralized. Depending on which part of Agoras it is, it might benefit from different marketing techniques. For speculators and gamblers, the referral marketing might work as they are risk takers and profit driven. On the other hand for miners and holders, there could be much more regulations involved so marketing of this would have to be done in a very different way.
Miners expect some predictable rewards over time, but they have what is called a variable interval schedule of rewards where they don't know how much time they have to mine but they know they will generally get X amount over variable range of time. Marketers (referrers) generally expect to be paid by commissions, and to create a commission type structure usually means a mix of variable ratio schedule with fixed intervals. This is to say marketers get paid based on how many new speculators or gamblers they bring in, and if gamblers are important for some speculative purpose of the network like pegging for example then just as exchanges use referral programs so too can Agoras for this.
What about developers? Developers have reputations and maybe reputation could help determine the reward rank of the developer so that developers with a very strong reputation get ranked higher and so are given higher fixed rewards, but you can also encourage developers using variable ratio schedule such as fixed rewards plus jackpot after the job is done. The size of the jackpot can be unknown but can be announced by the network to be within a certain range, and when the jackpot range gets big enough it can attract developers. The exact amount in the jackpot (reward pool) can remain encrypted and thus unknown but greater than a certain number as the Yao millionaire problem example shows.
Conclusion
- Game Theory is critical. If we get it right then the profit maps can be part of the marketing. The profit maps are the things earners can do to earn, gamblers/gamers can do to win. A profit map is like a graph, with paths which lead to different places on the map, where for example earners can become holders who then compound, or where jackpot winners can become gamblers + holders compounding etc. People generally want to see ways they can profit, survive, live, off the network, and the network seeks ways to improve itself.
- The two examples of models I gave were the vending machine and the slot machine. These are abstract models or mental models to help people think about the problem laterally. A vending machine is just a finite state machine, but a Rubik's cube or Sudoku is also a finite state machine. For example Sudoku can be represented as a list of lists forming a puzzle, but in technical terms it's a finite state machine as there is a finite amount of possible configurations, and the same is true for a Rubik's cube. A vending machine is a particular kind of finite state machine where you could for example select Sudoku from the menu or select the Rubik's cube from the menu. The vending machine is a good abstract model for thinking about smart contracts or for traditional software but the slot machine works better for thinking about speculation stickiness, or mining, or contests, or games.
- Trust is a result of having a reputation economy with the right game theory. That is to say the incentives if set up right will encourage participants to favor being trust worthy rather than being trustworthy. If the probability of trust being violated becomes low enough then you have trust but it is never guaranteed. Essentially what you can know is that for the potential earner, buyer, speculator or whatever class of participant, the rewards far outweigh the costs if they cooperate with the protocol. If they go against the protocol you can know the costs outweigh the rewards and so expect that the probability of this occurring will reduce.
References
- https://en.wikipedia.org/wiki/History_of_Microsoft#1985%E2%80%931994:_Windows_and_Office
- https://www.itemis.com/en/yakindu/state-machine/documentation/user-guide/overview_what_are_state_machines
- https://en.wikipedia.org/wiki/Game_theory
- https://www.casinosmash.com/features/how-to-use-probability-to-win-at-slots-2814.htm
You an economist? right?
Wrong. I don't consider myself an economist. I study what I need to study to learn what I need to learn as needed.
Thought you studied economy, nice job :P