Applications beyond your imagination in collateral mechanism and prediction market

Special thanks to Vlad Zamfir for his advice on prediction market content curation.

Over the past six years, people have been looking for creative applications of blockchain technology that they hope will become mainstream products. For cryptocurrencies, many of the applications are known - of course there is still the question of how it can maintain its edge as traditional payment systems continue to improve their efficiency. What about smart contracts and crypto 2.0? One approach is to compare traditional legal contracts today and think about whether the efficiency would be improved if smart technology was used. Another is the Peter Thiel-ian 0 to 1 approach: see if it is possible to use the tools to create a whole new field. So, let's examine this path and look at some underrated and interesting applications of smart contracts.

When Hashcash meets Proof of Stake

Although proof of work (POW) is well known today as a blockchain consensus algorithm, its initial application was not actually Bitcoin. The first major application of proof of work was Adam Back's Hashcash , a protocol to combat spam by making it very expensive to generate. Similarly, proof of stake (POS) is well known as a blockchain consensus algorithm, and its theory is based on the fact that proof of stake works in real life. In reality, proof of stake allows people to safely invest large amounts of economic resources, usually by spending them. Proof of stake can potentially be more efficient. This leads to an interesting question: can we use these advantages of proof of stake to create a more efficient version of Hashcash?

Let's clarify this first. The basic theory behind Hashcash is based on the idea that today's emails tend to fall into two categories:

1. Useful emails are emails that people spend a lot of effort to write and are valuable to read;

2. Useless emails are emails that people spend little effort to write and read with negative energy.

So, in theory, if each email costs a little extra to send, then useful emails will cost very little extra to send, and useless emails will become unprofitable. The cost comes in the form of electricity and computing power, which is used to solve a mathematical puzzle that can be quickly verified by the recipient before the email is displayed to them.

There are three problems with this attempt.

First, spammers could simply use a strategy of putting a little more effort into each message (like spending 5 seconds of human effort on each message to make it more appealing to readers or to get it through spam filters), but the bar for computational effort required to block most spam would be very high.

Second, spammers are highly specialized and have easy access to specially customized computer hardware that can solve these computational problems quickly and cheaply, so they only need one-tenth of a cent to do what the average user would cost 5 cents.

Third, the now famous checklist of problems with the popularly recommended solution showed how desperately we needed a system that could send mail freely.

So, this is the proof of stake option. When you send an email, you don’t have to solve an expensive computational puzzle, you just send a transaction to the blockchain, which creates a contract with some deposit in it as a security deposit. As part of the email, a private key is sent to the recipient, who can enter this private key into the contract to destroy the deposit (or donate it to a legitimate charity) if they wish. If the deposit is not destroyed within a certain number of days, it will be returned to the sender. Note that destroying the deposit does not benefit the recipient in any way - the only motivation is pure hatred. This gives us a clear contrast: the average cost of sending an email is low for the average person, because the recipient will only click "Report Spam" if the email is dangerous, which is rare, but the cost to spammers will be high - even the advantages of specialized custom hardware can't help.

People can see deposits up to $1, or even a gradual strategy: the sender can send any amount of deposits above a minimum, but the minimum threshold for the recipient to see the alert is an exact number. If it's only $0.10, it's just a message in the mailbox. If it's $1, there will be some phone alerts. If it's $500, their phone will be ringing at full volume, overriding all other settings - but the sender better be ready to pay if the recipient is worth the sender's unsolicited interruption.

We can also make advanced versions of this strategy, for example, instead of sending a transaction to create a new deposit for each email, we can imagine a strategy where the sender sends many keys to destroy part of the security deposit, with signatures saying that the keys are valid, and then the recipient publishes these signatures (but not the keys) in a channel like a bulletin board, allowing them to be quickly copied and ensuring that a particular deposit is not oversubscribed by destroying the keys (a specific way to do this is to only consider signatures containing indexes 1 to N as valid, and add a rule that two signatures at the same index can destroy 10% of the deposit, so it is easy to confirm that the deposit has at most N signatures). This way the transaction load can be reduced to about one transaction per sender per year. In any case, smart contracts provide a lot of room for creative optimization of details.

Prediction Markets and Reddit

One of the biggest debates about online communities like Reddit is how much centralized moderation is appropriate. One view is that much of the Internet’s appeal comes from its egalitarian, decentralized nature, and the fact that no one party has superior power over another. Some people may have more influence than others, but (i) that’s a matter of degree, not category, and (ii) it’s ultimately about audience choice. Another view is that without centralized moderation, communities will inevitably collapse into unwanted mediocrity and noise. Ultimately, Eternal September, and having a small group of users ultimately in charge, is, as it is in many places, a necessary evil.

In reality, the community’s voting and election mechanism is very powerful, but there are benefits to the centralist view. Although a speech that a community does not need to see is eventually voted down, at least on Reddit, this process takes a little time, and the content is still on the front page for a period of one or two hours. In the voting and election framework, this is somewhat inevitable: if you can quickly remove content from the front page through intensive voting, it itself has become a censorship mechanism for minority voices. However, what if there is a third way to solve this problem? This way uses a management mechanism that we are familiar with: prediction markets.

Prediction markets are now often introduced, including by myself, as a governance mechanism that could one day be used to make decisions about important issues: whether we need to bail out the banks, hire or fire a key CEO, or make an important trade agreement. However, as a decision-making tool, they work better on small, insignificant issues, those involving hundreds of dollars or even 10 cents.

One design could imagine working as follows. Instead of simply voting on hypothetical PredictionReddit posts, upvoting or downvoting, one could place bets on those posts in a prediction market. The prediction test is initiated by a mandatory bet made by the person making the post that their post will be considered good, so that as people vote, upvotes and downvotes will cause the market price to fluctuate. 99% of the time, the market has no effect, except that some high-priced posts maintain an advantage in the interface; the remaining 1% of the time, the post will be submitted to the meta-moderation mechanism panel, where it will be determined whether the post is good or bad (or something in between), and then the participants in the prediction market will receive some compensation, depending on the accuracy of their predictions.

In theory, the meta-moderator panel could be very large; every single participant could potentially be included, creating an effective anti-sycophancy mechanism. Even the SchellingCoin oracle could be used. It is not necessary to discard 99% of the market; one could use a model where all markets are processed, but only a small fraction of the meta-moderator panel looks at everyone's posts; the number of people would need to be large enough that it is virtually impossible for them to collude to trade inside the prediction market. Another option would be to scale the size of the meta-moderator and the probability of using it in proportion to the size of the market, so that the posts that get the most attention are the riskiest ones. In all cases, the special meaning of combining Reddit and cryptocurrencies looks much more promising than simply using 3 cents to comfort the relatives of the deceased.

In theory, both models can be extended: imagine ads that are more annoying to the viewer, but more expensive; or decentralized search engines where anyone can plug in their own ranking algorithm, participate in prediction markets, and only earn money if the algorithm works. Oleg Andreev's 2-for-2 escrow could be enhanced with a reputation system, by setting up a prediction market for the probability that the escrow deposit will be destroyed or delayed. Remember, security deposits and prediction markets are effectively equivalent: a prediction market is a security deposit that anyone can challenge, demand more in response, and support the initial depositor, while a security deposit, on the other hand, is a prediction market where one party is forced to bet.

Perhaps this is, in large part, what Crypto 2.0 can offer: using incentives to extract information from each of us more intelligently, thus elevating the Internet from a simple information technology to an economic information technology, and likely rapidly improving efficiency, at least in certain areas of the digital economy. In any case, let's create the tools and see.

Original article: https://blog.ethereum.org/2015/11/24/applications-of-security-deposits-and-prediction-markets-you-might-not-have-thought-about/
By: Vitalik Buterin
Translator: gsj
Reward address: 1DHs7hZMdx6tgUHdqXiub3XrovrF6Pte7y
Source (translation): Babbitt Information (http://www.8btc.com/prediction-markets)