Chapter 5: How smart contracts could work

Posted on by

[Note: below is chapter 5 to Great Chain of Numbers]

Theory is grey

While they do sound neat in theory, as Dr. Faustus discovered, “theory is grey, life is green.”1 One problem with institutions is not that they do not follow rules but rather that there is no conceivable set of rules that could unambiguously cover all of their activities.  Thus, to minimize nebulous outcomes, it is imperative for a programmer or businessman to conduct the necessary research and gather all of the requirements needed within the design phase of a smart contract.  This will be challenging one of the reasons that few known decentralized autonomous organizations exist today is that they could likely face various vulnerabilities and exploits that prevent them from carrying out their duties.  In fact, in his presentation to the 2013 Turing Conference, Mike Hearn (a core Bitcoin developer) noted this point: implementing the theory is much more difficult than creating it.2 Thus while science fiction novels and movies tease our imaginations with seemingly intelligent AI agents, creating even simple forms of non-creative bots will be a tall order.

Time Clock and Log-in

Over the past century there have been multiple mechanisms used by employees to verify that they worked a particular shift at a particular location.  Depending on trust levels, an employee may only need to say hello to their boss, others may need to sign their name on a particular line in a notebook.  Others might need to use a “card” that is punched with a timestamp throughout the day (e.g., when an employee first walks into the office, at lunch, after lunch and to clock out at the end of the day).  And there are even other employers in the past decade that have installed tracking software on computers.  While it is easy to verify that an employee is logged into the network or that an employee is indeed sitting at their desk and superficially looking at the monitor, some employers want to know exactly what is happening on each machine.3 Thus after an employee logs into the system, the software can siphon all of the input metrics (e.g., website visits, keystrokes, files) that they create during the day or other programs that randomly takes snapshots of the screen to verify that an employee is not watching videos when they are supposed to be filling out TPS reports.4 There are ways that each of the older “analog” systems can be abused.  In the case of time sheet or even time card, a friend or colleague could be asked to stamp your card even if you do not go to work.  Yet with the advent of software or even network-based tracking, it is much more difficult, if not impossible to abuse an on-site computer without making the company aware that the software has been removed or the network has been hacked.

Again, the goal of smart contracts and smart property is not to intentionally build some kind of totalitarian panopticon, but rather to enable all parties to clearly codify their responsibilities, obligations and compensation.  As I described earlier in the chapter, marginalized individuals such as migrant workers in China have little recourse during contract disputes due to the household registration system (hukou).  And they have a lot to gain if their contracts are not only tamperproof but that they can also prove in some manner if they fulfilled the contractual obligations such as on-site time.

While some requirements will be more difficult to codify into a smart contract, one area of low-hanging fruit could be the time-honored clock “punch.”  There are several ways to do this with existing systems: by using an RFID badge or NFC chip inside a phone, the “clock” would just have to be connected to whatever mechanism and network is ultimately responsible for sending the affirmation signal to the smart contract or DAO that automatically pays them.  Another example would be to use biometric fingerprints or eye scanners to verify the employee is “clocking in” (or out) and then connect that system to the same mechanism mentioned in the previous example.  There are limitations however: for instance, if an employee or contractor gets a piece rate or must frequently switch sites throughout the day to different neighborhoods, campuses, or even cities.  Creating a tamper resistant mobile check-in device that replaces the immobile clock to keep track of the number of pieces could be a business opportunity in the future.5 In fact, through the microtransaction abilities of Bitcoin, users can send micropayments, signed with their digital key, to prove that they were in a particular hotspot for a particular amount of time.

Decentralized Autonomous Organization

As I described in the introduction, a DAO is a virtual AI agent capable of performing, fulfilling, and executing the tasks, actions, and functions normally conducted by managers and executives, such as paying bills, issuing dividends and even crowdfunding an IPO.678 This would be done in a trustless or quasi-trustless environment, the “balance of trustlessness” determined by the intention of the parties and the capabilities of the code.  By using a Turing-complete language integrated with a cryptoledger, a DAO is essentially a tamper-resistant or tamperproof entity, immune to many of the abuses and vulnerabilities that have been happening to brick-and-mortar organizations are today (e.g., burglaries, arson, unintentional exposure to proprietary documents).  Currently no real decentralized autonomous organization (also known as a decentralized autonomous corporation or autonomous agent) is known to actually exist on a cryptoledger, although there are payroll bots and various software-based HR tools out on the market that integrate at the edges (BitPay).9

Some analysts claim that Bitcoin itself is a DAO because all of the users technically must submit a digital key which counts as some kind of voting mechanism, shareholders (miners) receive direct compensation for their work (seigniorage) – and there is no administrative overhead per se.1011 Yet, since development and direction of the Bitcoin protocol itself is not handled by direct ”votes” it is thus more akin to a proto-DAO.12

But voting and separate personality does not a company make.  Just like the cargo cult on Vanuatu dressed up like soldiers with the belief that air cargo planes would return with wartime goods, implementing voting into a cryptoprotocol and assuming this will create a company is a fairly superficial understanding of a corporation.13 Because of how development has come under the purview of the Bitcoin Foundation, the current Bitcoin ecosystem is a blend between “shareholder” and “stakeholder” system.14 This has potentially destabilizing issues in the long-term: fiduciary responsibility boundaries are fuzzy due in part to how it is funded (sponsorships) and how the organization wants to be perceived from the outside.  Furthermore, like any initiative there is the possibility that the network could be abandoned by users; a company cannot function without shareholder input.  This is not to say that there should not be a foundation (or many foundations) or even that a foundation could not receive money from outside sources or that users will abandon the project and network – rather, that because there is no direct voting process by bitcoin holders (like in a real corporation), the decision making process of the actual direction of the protocol itself is not an example of a DAO.

Last fall privacy advocates objected to a new “Coin Validation” project (whitelisting of bitcoins) and subsequently started the Dark Wallet and Zerocoin projects in an effort to move development one direction.15 While core developers have differing views, there have been no direct votes with digital signatures by bitcoin holders in this process.16 In fact, in the face of the new Coin Validation route that foundation members discussed, Roger Ver’s Blockchain.info promoted Shared Coin (developed by Gregory Maxwell, a Bitcoin developer) as a way to work around potential white and blacklisting.171819 This is not an endorsement of any proposal, but rather serves as an example of how a DAO could be used to mollify a set of actions.

Putting the DAC into DACP

Vertical institutions traditionally have created hierarchies in which intelligence and decision making is conducted at the top and automation functions based on guidance from human inputs.  An illustration of this phenomenon is legacy companies that arbitrarily trim divisions to meet certain metrics and consequently often cut at the edges of the network.  The lower echelons of departments in this case are sometimes viewed as replaceable or some simply lack the political capital (guanxi) that other departments may have had.  However, this dynamic all changed when Bitcoin introduced the idea of autonomous distributed consensus, automation at the center of the network and intelligence on the edges.

Unlike in a legacy company where decision-making authority is concentrated at the executive level, in a Decentralized Autonomous Consensus Platform (DACP), the decision-making authority is part automated, in that it has specific rules that are followed without possibility of deviation from expected form, and conversely human interaction and bias is limited to the edges.20 In such a model, power and authority, rather than being collected at the top, is spread to the edges of the network by allowing key holders (or VoiceHolders) to influence the decision-making and priorities of the DACP proportionally with their preapproved voting rights (e.g., when setting up a firm, a voting structure is put in place usually based on the amount of equity or shares an individual has).21

The legal liability and responsibility of a DACP or DAC still trace back to the key holders who sign their digital keys with a DACP which then calculates results based upon the prearranged voting proportionality.  For example, Bob’s Boutique requires that the allocation of special funds used by the DACP to hire contractors must be approved by a threshold of digital signatures.  If the threshold is unmet then the DACP does not release the funds to hire the contractors.

Software-based solutions used to calculate, authenticate and verify shareholder votes for many corporations and organizations already exist yet most still rely on a trusted 3rd parties and are susceptible to social engineering and man-in-the-middle attacks.  Thus, one opportunity for e-voting enterprises is to build consoles and virtual applications that utilize a cryptoledger, allowing members of organizations and institutions of all sizes to securely sign policy decisions.  To prevent internal takeovers and allow for quick dissolution (e.g., to manually reallocate assets), a self-termination clause could be programmatically designed within a DACP that could be triggered if enough shareholders submit signatures (or Voice) to a specific internal address within a specific timeframe (nLockTime).

In another real-world example: a DACP can be created as articulation of an assurance contract based upon a predesigned outcome, those who agree with the sentiment send funds to the DACP (the fundraising) and after a value-threshold is met, the DACP acts to bring about the desired result. Funds that are raised during this process are not releasable until a threshold (e.g., 51%) of those who put the value there in the first place, or those who purchased extra shares (e.g., giving larger voting pools) agree to both the need for the expenditure, and the final product being submitted for reimbursement.

Again, as mentioned in chapter 4, although this may sound futuristic, these autonomous platforms have no “artificial intelligence” at the top of the pyramid.  Where the capstone used to be the ultimate centralization of power, in the words of Adam Levine, “now it is only the nexus point for consensus from those participating further down the structure.”   It has the ability to spend funds, but only at the direction and authorization of the majority of shareholders.  Just like Bitcoin, DACs and DACPs are consensus driven, rules-based systems.  To be part of the system, according to Levine “is to follow the rules, so there can be no pre-mining, no individuals with privileged status at all.  Privilege is the antithesis of efficiency, and these structures seek efficiency above all things.”

Below is a rubric designed by Levine in his forthcoming paper to describe a hypothetical DACP assurance contract:

  1. DACP specification is proposed with Kickstarter Address collecting Ethereum/Bitcoin
  2. Received funds comprise development funds and initial DACP monetary base
  3. Kickstarter Address hits funding threshold and DACP Proposal Hub bounty is issued and rewarded by DACP consensus
  4. Proposals to develop DACP are created, and one or multiple are accepted
  5. Completed bounties are reviewed, and bounties are released by prearranged consensus.  DACs cannot integrate submitted bounty solutions until the winner has been paid.
  6. Once the platform is created and operational, DACP token holders can sell their tokens for Ethereum/Bitcoin at current market rate, hold it to speculate on the platform becoming more popular relative to the fixed number of DACP tokens, or exchange their DACP token with the DACP itself for DAC token as described above.

Experimental Cases

What a DAO could do is actually execute the contract based on pre-agreed to conditions.  If a digital signature counts as a vote, the only way to modify what a DAO would do is to get X amount of votes to approve some kind of execution process.  The specific amounts are hardcoded into the program beforehand and perhaps some are weighted differently.  To a limited extent, multisignature transactions, also known as m-of-n transactions (e.g., “joint bank account” “multisignature lotteries”), already work with Bitcoin itself, although again, you are limited to around 10,000 bytes, which would not be enough to fit hundreds of “votes.”22

Multisignature authorization of transactions is not a new concept as it has existed for hundreds of years in every corner of the globe.  This is done, as Szabo pointed out in chapter 3, to force conspiracy to take place in order for abuse to be undertaken.  That is to say, no single individual has the unilateral ability to abuse the treasury of an organization (or launch a ballistic missile).23 For example, using the Bitcoin protocol today as established by the built-in rules of Script (the name of the internal language), three parties could sign a contract which is programmed to release funds so as long as it receives the digital key of at least two of the parties.  As a consequence, this makes the Bitcoin protocol the legal system as it is impossible to use the tokens without the signatures.  Or in other words, if Bob operates a small company he may need to have 2-out-of-3 executives sign a document in order to release funds to pay for warehouse expansions.  With cryptocurrencies, the same idea applies wherein to move a ledger value (a bitcoin) to a different address, a smart contract or DAO that holds and controls “locked” tokens needs a predetermined amount (threshold) of digital signatures to release them.24 While it is not a DAO, Bits of Proof has developed software that provides this type of 2-out-of-3 reconciliation with a company, Bullion Bitcoin.25

In the future, a small auto-body company could create a DAO on the Ethereum ledger (or Litecoin, Bitcoin, etc.).  The company has five executives, each with a digital key needed to utilize and modify the cryptoledger.  Based on the company charter (and as specified in the smart contract or DAO), at least three of the five are required to use their keys in order for the tokens within a DAO to be used.  After a company meeting, an agreement is made to use the funds and three executives – Alice, Bob, and Carol – are asked to use their digital signatures (keys) to tell the DAO to release a certain amount of tokens.  Utilizing their smartphones (or any network connected device with an app tied into the ledger), they then submit their key and the funds are released.

This can scale up in the case of shareholders of a company.  Unfortunately as noted above, the current Bitcoin protocol has technical limitations that prohibit hundreds of digital signatures being sent to a specific address.  Yet other projects like Ethereum could potentially enable hundreds or thousands of signatures to be sent to a DAO.  This then could enable shareholders to vote on specific policies.  For example, if the board of a shoe manufacturer wants to expand production of a new running shoe that requires the use of tokens managed by a DAO, based upon pre-approved programmatic rules, they would need to bring this up for a shareholder vote.  A DAO could be preprogrammed to fulfill specific functions based on voter thresholds, like a majority or supermajority of votes (51%, 67%, etc.).  In this example, if there are 1,000 shareholders altogether, the DAO which was programmed with a 50.1% threshold, would only release the tokens if it received 501 digital signatures from all shareholders.

Peercover

In January I spoke with Jared Mimms who is working on Peercover, a startup that allows anyone to become their own decentralized insurance company.26  After months of work, they created one the first known smart contracts using a cryptoledger, interfacing with Ripple.27 According to him, “Peercover’s goal is to allow for a sandbox where people can chain smart contracts together and produce profit bearing assets (companies) without having to code. This mean the companies provide valuable services and are simple for people to use and join once companies are founded.  Peercover has developed a series of what they call “company types.” Each of these is really just an “algorithmic framework” for a company, including an “offer system” that allows founders to invest in companies by chaining 3rd party services to them to make them more attractive to join. Finally, a built-in trading system and soon to launch Simple Stock market allows founders to sell portions of their assets and investors to easily trade equity and reap automated or manual dividends.”  Mimms claims that Peercover is “the first true contract client in the space” which likely will increase competitive attitudes from other projects.28

With respect to smart contracts more broadly, Mimms says, “these types of instruments could provide a real opportunity for decentralized innovation.  Specifically, I saw how cryptocurrencies can allow for the automation of superfluous corporate functions.  And to accomplish this I began working with Peercover, where we can provide customers and entrepreneurs the ability to trade through gateways (via Ripple) without having to build and manage an entire backend.  Ripple has an open API that we use because currently it is the most efficient and robust at enabling  truly decentralized merchanting with low confirmation times compared with competing APIs that can take an hour per confirmation.”

Ripple Labs open-sourced the Ripple protocol last fall; the Ripple network has confirmation times between 5-15 seconds versus several minutes for blockchain-based ledgers.29

“For our first “smart contracts” we initially focused on peer-to-peer insurance companies – contracts – because of the new Obamacare mandates.  That is to say, there is a noticeable absence of insurance startups in healthcare and our platform makes it easy for companies to build their own custom solutions.  While we call them “companies” they are essentially a simple decentralized autonomous corporation (DAC).    Furthermore, one of our current plans is to integrate social networking functionality within Peercover to allow people (developers, customers, merchants) to talk to one another.  As a consequence, part of this process will require taking necessary steps to prevent fraud, thus we will verify people’s identities.  This may sound easy but as we have learned with working on various altcoin projects, if there is money involved some people will go to great lengths to commit fraud by forging and doctoring “official” photographs.”

The DAC claim is quite bold as no other team besides Invictus has announced any such development in a production environment.  And while legal compliance issues such as Know Your Customer (KYC) compliance have briefly been mentioned in passing; authentication has been a hurdle for other parts of the ecosystem, especially involving exchanges.30 According to several investors I spoke, maintaining KYC databases will likely become outsourced to firms that solely focus on this area of law.

Altcoins such as Dogecoin and altprotocols such as NXT surprised Mimms and his team this past year and he credits these two specifically for introducing a new marketing mechanism and potentially new platforms.  It is through these experiences that “we have learned to become adaptive and open to new cryptocoins and cryptoprotocols.  Because of our trial-by-fire experience, we can integrate with a new altcoin or altprotocol within a few nights whereupon we then provide users with a very flexible sandbox and drag-and-drop functionality to all users.  For example, if you own a bicycle repair shop you can create a customized contract that enables funding options, stock issuance, dividends and even discount management (e.g., 20% coupons to all users).   We are also the first company to actually create contracts that allow for accredited investors to create crowdfunding in compliance with SEC laws.  That is to say, instead of paying an investment bank like Goldman Sachs or J.P. Morgan to IPO your stock, you can do it yourself through a $200 kickstarter on our platform.  You can issue dividends and allow other people to hold shares.”

This crowdequity meme is also discussed in chapter 7 with examples from BankToTheFuture, JoinMyIPO and LTBcoin.  How the legal issues will be resolved in countries such as the United States is also an aspect to look into if your company is interested in this competitive space (e.g., allowing non-accredited investors to invest).31

Looking toward the future, Mimms says that Peercover has “also begun development towards using Watson-like functionality to provide fully autonomous customer service.  As a part of this effort, we have begun implementing tools such as a “tax” tab that will automate the amount of taxes to be withheld (e.g., a percent based on regional sales taxes that can be sent to specific cryptocurrency address).   While Ripple Charts users themselves are creating and executing contracts – which we build from, we have also partnered with BIPS, a large European-based payment solution provider in this space.”32)

Watson is a natural-language processing system developed by IBM and was popularized in a 2011 series of competitions on Jeopardy in which it beat two championship level human opponents.  IBM has subsequently improved on its abilities and plans to integrate the system in the healthcare industry.33 In addition, automated tax tools are another relatively “simple” area that several developers and investors mentioned that could be relatively easy for development and production.

Continuing, “While other platforms have noble goals, we think a decentralized 3rd party tools creates too much unnecessary complexity for end-users.”

Thus the takeaway message from Peercover’s experience seems to be, “in contrast, what we think is going to win is to have a sandbox-based platform that integrates fees where anyone can create and manage – with advanced interfaces – blindingly simple contracts.  For instance, we had one customer who raised $30,000 in two days with just $200 in kickstarter fees.  The technical backend, how it is done is not necessarily relevant to the minds of users who do not have time or the knowledge to fine tune the infrastructure.  And in the end, if your goal is to decentralize banking, keeping it simple is probably the number one issue developers and entrepreneurs should continually pay attention to.”

Subledger

The ability for a centralized platform to tap into decentralized processes is also being capitalized on with another project called Subledger.34 Subledger is an in-application accounting API that enables developers and businesses to integrate financial databases, including those based on cryptoprotocols into a double-entry real-time ledger analytics engine.

In February I spoke with Tom Mornini, co-founder at Subledger and according to him, “Applications make entries into Subledger for every transaction, in real-time if possible.  It’s then easy to share account records with the parties they represent, such as customers and vendors. That builds trust by eliminating the need for it, just like the blockchain does in cryptocurrencies.”  They have also refined segmentation so that customers have individual accounts for anything that needs to be tracked; we only aggregate during reporting.  Furthermore, the system never updates old entries and completely documents each transaction to maintain an audit trail.”

Continuing, “most people think accounting is about money.  While it is nearly universally used to track money but it’s really about tracking state changes of units of account which are not necessarily money.”

Furthermore, ignoring the time to close (e.g., the lag time between the close of a quarter and the close of the books) is a huge problem with currently deployed software. “The time to actionable information is critical for all companies.  The effort to audit is also greatly reduced, which also means less expensive, and auditing can now take place in real-time.  Auditors could verify a percentage of transactions every day, hour or minute and, essentially, continually attest to the accuracy of the information.”

He also sees at least one competitive advantage between Subledger (a trusted 3rd party) and DAOs, in his view, “the distributed autonomous organizations will be more expensive per-transaction because of the consensus overhead.  If there’s no counter party, there’s no reason to pay that overhead, which also requires the information therein to be public knowledge.  I’m not clear that a DAO would want its internal cost accounting to be shared publicly.  In some cases, yes, in other cases, perhaps not.”

Perhaps developers in this space can leverage a service like Subledger to provide a SaaS-based automatable system that integrates with an intranet-based cryptoledger as described later in chapter 8.  In the wake of Mt. Gox’s bankruptcy in February, which appears to have occurred in part due to a lack of internal accounts reconciliation practices and metrics, perhaps future exchanges could utilize a DAO or a CAO to provided quicker information to decision makers.

Where the Rubber Meets the Road

Currently it is difficult to foresee how the arbitration mechanism in a DAO would initially help anyone in China or other jurisdictions.  After all, who or what would enforce its decisions?  Or, if you used a DAO, what clauses could you include that hedge against the uncertainty of a potentially untrustworthy party?35 In terms of payment, fool proof clauses must be written into a contract that specify what exact channels or addresses the funds will go through and at what specific times.  While direct deposits are common on the mainland, it is not unheard of for unscrupulous employers to change bank accounts in an attempt to not pay debts.  Yet, cryptoledger-based escrow and bank providers may find new opportunities in this segment.36 And there are other options such as atomic-based transactions or atomicity in database parlance.

Michael Goldstein, founder of the Satoshi Nakamoto Institute, wrote a concise explanation of what an atomic transaction means:

Two parties agree to exchange one cryptocurrency for another, and the transaction is done in such a way that neither side can execute their portion of the trade without releasing funds to the other party. The trade either happens in its entirety, or not at all, which means nobody can walk away empty-handed. The worse possible outcome is that no trade occurs at all and everybody keeps what they had.37

You can substitute “cryptocurrency” with any kind of token (metacoin, colored coin or even a smart contract) that is capable of performing the same function.  Previous such atomic transactions have taken place in other systems such as with airline bookings.  A potential passenger must both pay for and reserve a seat or neither pay for nor reserve a seat.  A booking system will allow one option to occur and not a mix.

Using existing technology plus atomic transactions, there are several ways an employee and employer could resolve payment disputes.  In a small business between friends, family and other trusted parties, the formalized contract steps could be minimized.  A simple contract might look like this: working from home (it could be any arbitrary location) Bob builds a website for Alice and thereafter uses 0.0001 bitcoin (or litecoin, etc.) to generate a temporary token of arbitrary color, size or type but which represents a predetermined, pre-agreed amount of value – a temporary “labor” coin – and sends it along a cryptoledger to Alice.  Later that day, Alice looks at and approves of the website quality and subsequently sends Bob a token of predetermined value worth $500 (the exact amount is based upon a previously agreed to amount) and utilizes the same cryptoledger (although it does not necessarily have to).  Both tokens have a function called nLockTime built into them for twelve hours (these time values are arbitrary).  If both tokens are sent and received during that twelve hour time period, then the atomic-transfer takes place and both receive the other token.  Alice scraps the token she receives because it was a mere abstraction of the labor Bob provided (she can keep it for accounting purposes if she wants).  Bob on the other hand can then exchange his token to any fiat exchange, token exchange (Cryptsy, Bter), or perhaps even a merchant.

Again, this was simplified to illustrate how the atomic transaction works.  In this case, if one or both parties did not send their token in the allotted time – none of the tokens would be received by the intended parties.  Instead, the ledger would send it back to the originating wallet address.  For example, if Alice did not send a token, the next day Bob would wake up and see that he has not been paid and his “labor” token was sent back.  He can then talk to Alice to find out what the issue might be; after all, he would like to be paid for his labor.  In reality, as well as this example, it is possible for both Alice and Bob to use different cryptoledgers so long as there is some mechanism like a web exchange that has the ability to process both types of tokens.38

And so long as there are decimal units in a particular token, the logistics of sending value can be scaled up, for all practical purposes, near infinitely and potentially infinitely (assuming scalability issues can be overcome).  Even if all seven billion humans (plus DAOs) immediately began using one particular cryptochain that used just one specific base token (a bitcoin, ether, etc.), they could send fractional token sizes (e.g., 0.00001) to other parties.  Each individual (and DAO) could also include a secondary attribute in a “hash” or code snippet to identify what asset this token actually is meant to represent, such as cars, commodities, “labor,” and fiat (e.g., using metadata to turn 0.0001 BTC into a “blue” token or some other random attribute that acts as an abstraction to an specific asset).

Abstractions and Decimalization

Consequently because of this decimalization, the virtual economy should never run out of base tokens in the money supply.39 This is not inflationary, as no new base token is created that is not tied to some particular asset.  The underlying foundational token is still tied to the scarcity of the original money supply.  Furthermore, there are built-in anti-spam functions in existing cryptoledgers that require minimum transmission values, below which a transaction is not permitted along the network – this is known as the dust limit.40 And again, with Bitcoin, every 10 minutes 25 bitcoins are “created,” in Litecoin, every 2.5 minutes the same amount is created.  Other cryptochains have their own known, invariable money supply creation rates, but this is not important as the fundamental ideas are the same in that these tokens can be further subdivided into increasingly smaller decimal spaces and also given a second attribute to represent a different asset.  Similarly, additional DAO-based banks and escrow services could provide functions if atomic transactions are not agreed upon beforehand.

At a large enterprise for instance, Bob, a graphic artist, arrives at Adobe (his employer) and logs in with an RFID badge at the front door of the office.  The clock sends an encrypted signature to an HR DAO run by Carol’s independent escrow that creates a timestamped ledger entry on the Bitcoin network (or Dogecoin, etc.).  Bob’s computer is also fitted with software that can monitor his inputs, which are stored on Adobe’s SAN (while this verification role is redundant, Bob could also later point to the information gleaned as proof that he worked).  After completing his assignments Bob again, clocks out with his RFID badge which sends another encrypted signature that generates a token that represents one day worked.  The color or type or size of token is irrelevant as it is merely a representation of an agreed upon completed condition.  Alice, his supervisor can later send her own signature of approval (or disapproval), which is then sent to Carol’s DAO.  If approved, the token could be released and sent to another independent DAO, Dan’s bank, which stores tokens held in escrow on behalf of Adobe (e.g., n-of-m).  Once both tokens are received, the transaction triggers a time-based predetermined, pre-agreed settlement clause between Carol’s escrow services and Dan’s bank whereupon Carol sends Dan’s bank the “labor” token (which could be discarded or held for accounting purposes) and Dan’s bank sends a pre-agreed token (e.g., a bitcoin worth $500) to a prearranged wallet address that Bob controls with his private key (e.g., his bank account).  Dan’s bank could also send the token through other DAOs, this was just one illustration.

As you can tell, this type of system could be used with any amount of time lock, including years – hence the long-term potential uses of managing trust funds and the execution of wills.  For example, Bob has $1,000 and would like to give it to Alice, his 1-year-old baby daughter, when she turns 21.  Bob has several choices.  He can immediately exchange the $1,000 in fiat for a token (e.g., bitcoin) and place it in escrow.  He could deposit the fiat into a bank and fill out a smart contract with the bank that provides a time-based disbursement condition (e.g. in 20 years, spend the $1,000 and purchase an equivalent amount of bitcoin and send it to Alice).  He can deposit the fiat into a bank but then create a smart contract-based financial instrument with a fraction of a bitcoin (0.0001) which may cost him a few dollars now and send the smart contract to a DAO bank where it sits until a specific date is triggered.  He could also simply exchange $1,000 fiat for a bitcoin token today and leave it on the cryptoledger using an “external state contract” because the nLockTime function is already built into the protocol and the token will automatically go to a prespecified address at a specified time (e.g., in 20 years it will be moved to an address controlled by Alice or Bob).41 However, Bob should also be aware that if he signs and broadcasts the transaction far into the future there is a chance that some nodes may choose to drop the transaction in the memory pool.  If he uses an escrow, he can also create a smart contract that lays out the specific conditions, the terms to which a token is allowed to be sent to a pre-specified address (perhaps Alice has her own address, or maybe she is given access to his in the event he dies).

Another way to handle an inheritance with the existing blockchain is through an entity called an “oracle;” an autonomous 3rd party agent.  In his 2012 presentation Mike Hearn described an independent, trusted oracle system that is set up to monitor the obituaries section of a government agency or newspaper, whereupon it can relay identities and information to a contract on a cryptoledger.  Or in other words, the oracle listens and uses that data to sign a multi-signature contract and in order for the contract to release funds to the beneficiaries, it needs a signature from the oracle.  The contract had previously been signed by the original trustees who require a signature from the oracle to release funds to predetermined beneficiaries.42 Last year Michael Goldstein described another oracle involving a sports bet: Bob bet that team A would win and Alice bet the other team would win.43 An oracle holds the deciding key to a contract that says if team A wins, Bob receives the funds (bitcoins) and if team B wins, Alice receives the funds.  The parties write the contract noting how the transaction should proceed in the event of disputes or potential ties.  Then after the event is over an oracle signs it removing the middleman.  Unlike ordinary legal disputes involving nuances and grey areas, sports betting is an objective, idealized scenario because there is no grey area as all that an oracle would have to do is have access to an ESPN data feed.

Mitigating Abuse

Other near-term uses within a cryptoledger are loyalty programs, merchant reward programs and “Frequentfliertokens” from Alice Airlines which could help prevent and mitigate the risks involved in travel hacking (e.g., getting frequent-flier miles without flying).44 For example, United Airlines frequent-flier miles were downgraded effective February 1, 2014, due to rampant inflation caused by a combination of website vulnerability exploits and quick scheduling changes by users.45

Instead, Alice Airlines could offload the auditing, storage and transportation of rewards and utilize the “contract” system of a cryptoledger by using an arbitrary amount of a token (0.01 BTC), creating a “contract” that defines a set amount of mileage (which itself will likely have some predefined expatriation dates).  Assuming that flyers are using cryptocurrency wallets and provide the airline with their wallet addresses, the users will be able to receive the mileage amount in their wallets.46 In turn the users can sell and trade the reward tokens by sending a specified amount to Alice Airlines.

Other institutions can use a smart contract to issue and track its own customer loyalty program rewards.47 For instance, in 2005, Subway ended its sub club stamp program whereby a customer would receive a couple of stamps (stickers) for certain purchases.  When a customer collected a certain threshold of stamps, he or she was eligible to receive free food (chips, drinks, sandwiches).  Yet, a number of customers found a way to game the system by buying and selling entire reams of stamps on eBay, creating massive stamp inflation costing the parent company an unspecified amount in losses.48

Thus, coupons are another ripe area for development.  According to NCH Marketing, “Consumer Packaged Goods (CPG) manufacturers distributed 305 billion coupons in 2012, the same quantity as the year prior. […] total redemption for 2012 fell 17% to 2.9 billion coupons, saving CPG companies a substantial $800 million in face value discounts.”49  While this may seem like a mundane area, consider that by 2016 Juniper Research predicts that “the total redemption value of mobile coupons will exceed $43 billion globally” because coupons are increasingly delivered by mobile apps.50 For perspective, 48% of adult internet users in the United States redeemed a digital coupon for shopping in 2012.51 A company providing coupons or discounts could create a DAO to manage these redemption contracts (e.g., a type of time-locked token), which will not only reduce the logistical overhead but also prevent coupon abuse and fraud (e.g., double-spending).  According to the US postal inspector Roberta Williams, “for every coupon successfully counterfeited, it costs the manufacturer $1 million.”52 Initially these fake coupons are scanable but the coupon inflation ultimately forces manufacturers to redeem more than they had intended.  Furthermore, the Coupon Information Corporation (CIC) estimates that coupon scams create losses of $300 million to $600 million a year – and that these costs end up getting passed onto consumers.53 Yet, if there is one function that the algorithms governing Bitcoin money supply have proven adept at, it is preventing inflation.

While not directly related to fraud prevention, one real world case-study within this overall segment began in February 2014.  PointsHound, a site that rewards travel reservations with frequent flier miles and hotel points, announced that it had begun using bitcoins in its payout system.  If a user selects the bitcoin payout option, PointsHound calculates the reward based on the market price listed on Coinbase and then sends the amount to the user’s wallet.  According to cofounder Pete Van Dorn, “We carve out a portion of the commission to give back to the customer in the currency of your choice. It might be 5,000 miles, it might be Bitcoin.”54

The Tao of DAO

In developed regions, market participants are familiar with computer software that uses, runs, manages, and executes nearly all of the financial instruments on electronic stock exchanges – and how there are various clauses written into them to hedge against (or prevent, or in case of) some type of counterparty risk.  Below we will look at how that functionality can be designed into a smart contract with a normal contract at a normal job or even one that a Chinese migrant worker may do.

One way this might work:  Bob, an employee, would use a digital key to sign a smart contract with his boss Alice, who also uses a digital key to sign it.55 Within the contract will be a number of provisions and stipulations regarding payment time periods and clauses that hedge against the possibility that one party does not fulfill his or her end of the bargain.  Perhaps there will be a clause that says how payment will actually take place: through an escrow service (BTCrow), through bank X, through address Y, or a mix of different options.  This contract could be stored on a public decentralized cryptoledger (e.g., Bitcoin, Ripple).56 If stored on a cryptoledger it is tamper proof and forge proof as it sits there immune from 3rd party interference.  Again, while most people think of Bitcoin as a currency tracking tool, in arithmetic terms it is more akin to a database that can be used to track any particular dataset (e.g., a bitcoin) as long as it fits within the technical limitations.  It just so happens that the sole data this past four years has been for one particular “token” as represented by an integer on the ledger (i.e., bitcoin).57

While there is a way to change the way the DAO could operate by convincing the rest of those with votes to modify it with their private keys, the original contract would still be left in public view and untampered with.  What could happen is that contract itself would have an nLockTime (time-based) clause or condition that after X amount of time, if certain conditions are not met (for example, payment) then it would follow some predefined termination clauses.  Perhaps it would send itself to a predefined arbiter or escrow DAO.  While it is doubtful that smart contracts will solve all of the problems on the edges of a network (e.g. brick-and-mortar infrastructure), it will prevent tampering with the actual contract itself thereby protecting employees (and employers) from trusted 3rd party risks such as fraud.

So in a nutshell, ignoring other aspects of asset management, the following scenario could take place:

Bob digitally signs a smart contract with Alice stipulating various expectations, terms of compensation, etc.  This contract stipulates that payment will go through various channels each month, however if there is a breach of contract it will end up with Cathy’s escrow service (which itself could be an independent DAO).  In fact, there will very likely be several virtual escrow services that need to maintain a good, honest reputation to do business (just as they do today).  Furthermore, there will likely be a dispute-mediation clause regarding independent arbitration if all else fails (just like today).58 By “all else fails,” I mean there will be a time-based trigger: if neither Bob nor Alice re-sign clause B, C or D by a specific time in the contract located on the ledger, the contract is sent to Dan the arbiter or Eve at the public court.  Dan could, like an independent escrow service, be chosen from a list of known reputable arbiters who face similar market conditions to provide unbiased service (net-ARB is a type of service like this).59

These types of default-based relationships and contractual stipulations take place today.  While it may be difficult to initially “codify” them into software, it is likely just a matter of time: last year Coinsigner became the first cryptocurrency-focused dispute resolution service using multi-signature transactions.60 In fact, the barriers to entry are low enough that individuals can create independent mediation resolution practice to provide objective, unbiased fair decisions and any three people can employ it, across borders.  Whether these systems will be commercially viable for enterprises of significant scale is, however, as mentioned several times in this manuscript, very much an open question.

If stricter capital controls and regulations on cryptocurrencies are enacted in China (or elsewhere), by using a couple different “colored” coin chains (or other ledger contracts), Bob from Beijing could potentially transfer assets worth X amount of money to Alice from Anhui instead of X amount of money itself.61 This could create a sort of advanced barter system which may not be as efficient in terms of actually using a cryptocurrency as a medium of exchange but it could help those in an informal economy qualify and quantify asset value and clear up some of the confusion around contracts and property ownership.62 Yet how those contracts will be enforced is also an issue that will likely fill volumes, as any institution providing such service will be a target policy oversight, just like the exchanges are now.

At the same time, there are some uncertainties and legal risks which will vary from one jurisdiction to the next.  In China, it is hard to speculate how the various townships, counties, municipalities, provinces and the central government itself will recognize this type of ledger-based asset management.  Twenty years ago, most Western commentators believed that the internet would empower the average Chinese resident to maneuver around censorship, but the Great Firewall has proven very capable of stemming the flow of all information.63 While it would be difficult for them to block such decentralized peer-to-peer activity perhaps each government layer will instead want a small piece of the transaction and only recognize smart contracts that go through specific government-run DAOs or corporate custodial escrow and arbitration services with a contractual nexus to real assets and a national system of law and enforcement, as suggested by Preston Byrne in chapter 2.

Share the post "Chapter 5: How smart contracts could work"

  • Facebook
  • X
  1. This is from Goethe’s Faust.  The actual quote is, “grey, dear friend, is all theory, and green the golden tree of life.” [↩]
  2. See Mike Hearn, Bitcoin Developer – Turing Festival 2013 video and Bitcoin Developer Mike Hearn and Amex VP Michael Barrett Join Circle Team from CoinDesk [↩]
  3. There is a keyboard shortcut and browser extensions called a ‘boss button’ which quickly switches the screen to hide certain programs (like a computer game or video). [↩]
  4. Companies can receive analytics that provide reports on all file and network.  Some examples of such software are Ultra VNC, eBlaster and Screenshot Monitor.  See Are You Being Monitored At Work? by Becky Worley [↩]
  5. Teaming up with “check-in” providers such as FourSquare could be a method. [↩]
  6. Gregory Maxwell uses the term ‘agent’ in his StorJ proposal; see StorJ, and Bitcoin autonomous agents [↩]
  7. BitcoinStarter and CoinFunder are two current services in the crypto crowdfunding space. [↩]
  8. It is currently unclear what the arbitrary distinction between an “advanced” smart contract and a barebones DAO lies.  Both use a blockchain to conduct and manage their organizational operations.  Furthermore, to use the TCP/IP and SMTP analogy it is unclear at this time whether Bitcoin is merely one type of crypto app (like SMTP) or if it is more general purpose – a foundation – like TCP/IP is.  The ’2.0’ projects in the broadest sense (like Mastercoin, Colored Coins, Invictus, Ethereum, etc.) are an attempt to create a more general platform more akin to TCP/IP that other services are built on top of. [↩]
  9. See BitPay, Coinbase and Bootstrapping A Decentralized Autonomous Corporation: Part I by Vitalik Buterin [↩]
  10. Bitcoin and the Three Laws of Robotics by Stan Larimer:

    Bitcoins can be viewed as a small “share” of the total market cap of the Bitcoin “corporation”.   The “mining” services that validate transactions and secure the network are paid for in new bitcoins that slowly dilute the “stock” as the corporation’s market cap ebbs and flows.  You can generally trade your shares for other currencies, goods, and services.  Operating rules for the corporation cannot be changed unless a majority of stakeholders vote for them by switching to another version of the software.  Interestingly, it is not the holders of existing shares that get to make this decision, but only those “employees” who are contributing their computer resources (mining bots) to run the company.

    Nothing says a corporation can’t be structured to distribute voting rights this way, and that’s exactly what Bitcoin has done.  Shareholders get equity growth.  Employees get voting rights.  All “revenue” is paid to the employees as compensation for their work.  There are no profits. [↩]

  11. An early concept of a larger voting-based system built on a DAO is the Bitcongress Foundation.  Furthermore, David Johnston of the Mastercoin Foundation articulated this same software development centralization problem in a January 24, 2014 interview, episode 80 – Beyond Bitcoin Uncut from Let’s Talk Bitcoin. See also DAC Index [↩]
  12. Vitalik Buterin labels it a prototype, stating:

    As Let’s Talk Bitcoin’s Daniel Larmier pointed out in his own exploration on this concept, in a sense Bitcoin itself can be thought of as a very early prototype of exactly such a thing. Bitcoin has 21 million shares, and these shares are owned by what can be considered Bitcoin’s shareholders. It has employees, and it has a protocol for paying them: 25 BTC to one random member of the workforce roughly every ten minutes. It even has its own marketing department, to a large extent made up of the shareholders themselves. However, it is also very limited. It knows almost nothing about the world except for the current time, it has no way of changing any aspect of its function aside from the difficulty, and it does not actually do anything per se; it simply exists, and leaves it up to the world to recognize it. The question is: can we do better? [↩]

  13. Richard Feynman first popularized this superficial hand-waving phrase 40-years ago through his memorable lecture, Cargo Cult Science.  The name is derived from the actions of a South Pacific tribe located on the island of Tanna in Vanuatu.  See In John They Trust from Smithsonian. [↩]
  14. See The Shareholder vs. Stakeholder Debate reconsidered by Rüdiger W. Waldkirch and How to Bureaucratize the Corporate World by Ben O’Neill [↩]
  15. See Sanitizing Bitcoin: This Company Wants To Track ‘Clean’ Bitcoin Accounts from Forbes and Coin Validation misunderstands fungibility and could destroy bitcoin by Adam Back.  Technically Zerocoin was already in development months before the Coin Validation announcement; at the end of the year Dark Wallet held a successful crowdfunding campaign. [↩]
  16. While the miners could collectively fork and begin hashing a modified Bitcoin that integrated with Zerocoin, they have yet to do so for a variety of reasons, namely the $1 billion in capital investment (hardware) that would have to be written down because Zerocoin uses a new ledger and proof-of-work.  See Anti-Theft Bitcoin Tracking Proposals Divide Bitcoin Community from CoinDesk, Bitcoin Anonymity Upgrade Zerocoin To Become An Independent Cryptocurrency from Forbes  and Hopkins researchers are creating an alternative to Bitcoin from The Baltimore Sun [↩]
  17. Shared Coin was originally called CoinJoin, see this tweet from Blockchain.info.  In addition there is a difference between on-chain (e.g., Blockchain.info) and off-chain wallets (CoinBase and Circle).  See Roger Ver on Blockchain’s Past, Present and Future from CoinDesk [↩]
  18. While the project is still in its early stages, the Ethereum blockchain will unlikely include the anonymity features of Zerocoin.  Rather, there may be ways to create smart contracts and DAOs which can provide some level of anonymity (e.g., shell companies). [↩]
  19. While these decisions provoke strong opinions and feelings, forks are also a potential as well.  Several “next generation” platforms may be compelling to some niches because of potential DAO functionality that could in turn use a contract or DAO to create ‘holding firms’ or even ‘shell companies’ (though obviously it is still on paper and has not been made).  Yet, even if something like Ethereum worked as stated and the Bitcoin development team coded in significant protocol and proof-of-work (PoW) changes, it is unlikely you would get even a plurality of Bitcoin ASIC miners, let alone 90% to agree with moving to a new PoW algorithm because that would make their capital investments worth exactly zero (because a Bitcoin ASIC is tuned to one particular PoW, SHA256d).  One recent estimate suggests that there is roughly $1 billion invested in existing hardware for mining globally including ASIC R&D.  See The Bitcoin-Mining Arms Race Heats Up from Bloomberg Businessweek.  Again, while “forking” comes up in conversations, ultimately the value is not necessarily the software code itself, but the infrastructure and mind-share behind it (the ecosystem). [↩]
  20. Adam Levine coined this acronym for descriptive purposes.  Portions of his upcoming essay “Application Specific, Autonomous, Self Boot-Strapping Consensus Platforms (And the DACs that live on them)” are rephrased and reprinted with his permission. [↩]
  21. This term is used to describe anyone controlling more than a certain amount of tokens that interface with a DAC(P).  The concepts of “exit” and “voice” were described in Exit, Voice, and Loyalty: Responses to Decline in Firms, Organizations, and States by Albert Hirschman.  These ideas have gained new prominence due in part to the decentralizing abilities and functions created by the software community.  See Software Is Reorganizing the World and Silicon Valley’s Ultimate Exit (slides) by Balaji Srinivasan [↩]
  22. See Why are m-of-n transactions not used today? from StackExchange, What are multi-signature transactions? from StackExchange, Bootstrapping A Decentralized Autonomous Corporation: Part I by Vitalik Buterin and the Ethereum whitepaper. [↩]
  23. This is called two-factor authentication (2FA) or two-man rule.  See also Shamir’s Secret Sharing. [↩]
  24. For a technical overview of how multisig works, I recommend watching a video explanation by Andreas Antonopolous (Taariq Lewis put together this slide deck of Andreas’ notes). [↩]
  25. Bits of Proof and Bullion Bitcoin to Launch Gold-Bitcoin Exchange from CoinDesk [↩]
  26. Peercover [↩]
  27. The team is also involved in a new broadcast marketing initiative at OpenXRPTalk [↩]
  28. Personal correspondence, January 31, 2014 [↩]
  29. Ripple is officially open-source! [↩]
  30. Compliance Program Design Presentation at First Virtual Currencies Compliance Conference in NYC by Juan Llanos [↩]
  31. See Accredited investor and the JOBS Act [↩]
  32. BIPS (Bitcoin Internet Payment System [↩]
  33. Doctors seek help on cancer treatment from IBM supercomputer from Reuters [↩]
  34. Subledger [↩]
  35. If the “clawbacks” over the past decade are any indication, local townships and provinces will likely be a hurdle until they see the utility such ledgers could provide their own administrations.  See China’s Turn Against Law by Carl F. Minzner [↩]
  36. See Example 2: Escrow and dispute mediation [↩]
  37. Lex Cryptographia and Satoshi Nakamoto Institute [↩]
  38. These sites already exist: BTC-e, OKCoin, Cryptsy and Bter are among the largest multi-token processors. [↩]
  39. Bitcoin has 8 decimal places, the last of which is called a satoshi.  In Ethereum, the last digit is called a wei.  Note: in all examples, each user uses a cryptoledger as the mechanism for transport and audit. [↩]
  40. The dust limit has changed over the years and was implemented to prevent transaction spam (e.g., tens of thousands of transactions each amounting to 0.00000001 BTC).  The current limit is around 5460 satoshi.  See DustTransactions and What’s the minimum transaction with bitcoin? from StackExchange [↩]
  41. Using external state [↩]
  42. The trustees do not necessarily need to be humans, as a DAO or Digital Oracle could technically act as a party and signatory authority.  For example, see the whitepaper Securing wallets by integrating a third-party Oracle from CryptoCorp [↩]
  43. Michael Goldstein Explains How The Bitcoin Block Chain Enables Smart Property from Newfination [↩]
  44. The Ultimate Travel Hacking Guide from Lifehacker and How to Be a Travel Hacker by Nomadic Matt [↩]
  45. Recap of United’s Downgrades: Award Charts, ExpertFlyer and Meals from Frequently Flying [↩]
  46. It could simply be a hash of an embedded URL that sends you to a screen on Airline Alice with the actual amounts along with the Terms of Service.  Colored Coins have this potential capability as do other projects like Ethereum. [↩]
  47. This is not to say that a company needs to build and maintain its own cryptoledger for a rewards program.  For example, assuming that Cocacolacoin is not using the Ethereum blockchain (or Bitcoin) but rather uses its own independent PoW blockchain, it may be hard to incentivize network hashrate which creates network security (which prevents a 51% attack).  That is to say, instead of trying to incentivize Bob the Miner to exchange hashrate for Coca-cola swag only, Coca-cola could simply use a common, independent cryptoledger (like Bitcoin). [↩]
  48. Fraud Sinks Subway’s Sub Club from Wired [↩]
  49. CPG Coupons: U.S. Market Analysis from NCH Marketing [↩]
  50. Mobile Coupon Redemption Values to Exceed $43bn globally by 2016, Driven by Better Targeting and Mobile Apps by Juniper Research [↩]
  51. Leader In Fast-Growing Digital Coupon Industry Sets Debut from Investors Business Daily [↩]
  52. New coupon scam is costing U.S. companies millions of dollars from Fox6 [↩]
  53. Start Your Own Online Coupon Or Daily Deal Business by Richard Mintzer [↩]
  54. Rewards Program Tries Bitcoin from The New York Times [↩]
  55. Public/private digital key [↩]
  56. BTCrow [↩]
  57. One of the primary reasons this was the case is because Satoshi Nakamoto intentionally created Bitcoin for that purpose, hence the full name of the paper “A peer-to-peer electronic cash system” – the first section of the whitepaper discusses the problems people have with paying for things online; it was not a manifesto. [↩]
  58. See Example 2: Escrow and dispute mediation [↩]
  59. internet-ARBitration [↩]
  60. A Decentralized Bitcoin Exchange Process Dreamed up and Executed from Coinsigner [↩]
  61. See Colored Coins project and Colored Coins: NYDFS Reviews Ways To Transfer Ownership With Bitcoins from International Business Times [↩]
  62. Many of these “crypto exchange” ideas trace themselves back more than 20 years both in academic literature (Nick Szabo) and in science-fiction (Neal Stephenson).  In fact, Stephenson wrote three novels in the 1990s which include crypto-based themes as an integral part of their plots (not cryptoledgers or cryptocurrencies, neither of which were foreseen).  These are Cryptonomicon, Snow Crash and The Diamond Age.  Prior to these publications, one non-fiction document that is historically seen as significant in the development of anonymous digital currencies and electronic privacy is The Cyphernomicon by Timothy May. [↩]
  63. The Great Firewall (防火长城) is an ongoing multi-decade project by several Chinese governmental institutions to filter and block undesired information from the mainland.  The GFW is very effective for the most part; without a VPN, I was directly impacted every day for 5 years.  I discuss this in Chapter 20 in Great Wall of Numbers.  See also The Master Switch by Tim Wu [↩]

Leave a Reply

Your email address will not be published. Required fields are marked *