The Problem with Smart Contracts
A lot has been mentioned about the benefits of Smart Contracts. Focus is always placed on the power of being able to automate contractual agreements by leveraging the Blockchain. Very little is mentioned about the mechanics of how the real world elements represented in the smart contracts, actually interact with the blockchain ie how does the smart contract really know whether all the conditions of a contract has been achieved in the real world.
Though smart contracts can be trustless, the inputs to the smart contracts aren’t. Furthermore the execution of the outputs of a smart contract rely on a real world agents to reliably execute the said terms of a smart contract. This presents a real concern in that a lot of blockchain applications will be deployed with the assumption that they’ve solved the trust issue in their respective domain, but in fact a lot of room still remains for centralised control both on the inputs and outputs of a smart contracts life cycle.
A classic example is the blockchain real estate use case. It’s all well and fine to encode property ownership into the blockchain, and allow property owners the certainty of their ownership claim to be in an incorruptible, decentralised and censorship resistant “database”. But the mechanism by which the smart contract establishes who has a valid claim on a piece of property, or how it determines and logs a change in a properties valuation, are still very much centralised via our legal systems. Essentially, all we have today is a way of recoding what we want to record in a trustless and censorship resistant manner, but we still have not solved the problem of ensuring that the process of establishing if an ownership claim is itself valid prior to being entered in the in the blockchain.
Thus the migration of information from the World of Atoms to the World of Bits where the blockchain and smart contracts reside is a core infrastructure challenge which needs to be resolved in order for smart contracts to truly become a viable replacement for a lot of the systems we rely on today. Until that happens care needs to be taken to monitor this failure mode before entrusting a smart contract with more power than it ought to have given the said constraints.
A natural response to this is to instead focus on the incremental improvements that smart contract bring over the current state of things. In as much as the inputs to the ledger can be still be centrally controlled, there’s still a lot of value in being able to automate contractual conditions, particularly when the results do not require a transfer of information from the world of atom to bits.
This is particularly true for a lot of financial products and services, which primarily can all be initiated and settled in the world of bits. One can extend this benefit to any situation which does not require much interaction with the world of atoms.
The legal system is another major source of friction, as it encapsulates all aspects of the how we interact with each other, for a large part, it’s the real competitor which smart contracts are geared to replace. Given that our legal system is still largely un-automated in that, there are still centralised around professional bodies and councils, who are legally empowered to hold a monopoly on a lot of the transactions we frequently engage in.
This digital migration of our legal system from atoms to bits, might arguable never happen, given that a lot relies on complex moral, historical and philosophical elements which maybe an Artificial General Intelligent agent will be able to potentially automate away from us, in a less conflicting manner than possible today. This is not to say that incremental improvements can not be made and appreciated, as smart contracts can be used today to remove a lot in efficiencies, but to themselves become the source of legal legitimacy, a major shift is needed.
A massive philosophical and political shift will need to occur to transfer legal trust from institutional legitimacy to algorithmic legitimacy via smart contracts. This shift will need more than then the failures of institutions to ride on, additionally a solid case needs to be made for what is essentially a Dictatorship of Algorithms versus the ideal of the existence of model human beings which we can elect and trust to do the things we cant ourselves. In Plato’s terms, it would be shifting from a rulership of Philosopher Kings to a Rulership of Algorithms. The two don’t have to be mutually exclusive, as common sense today would probably dictate a compromise of the two.
All that being said, there’s a lot of blockchain companies that have identified this problem and are working on building out solutions to allow smart contacts to access the real world. One notably is Chainlink, which takes the approach of building a network of Oracles Nodes which attempt to act as a source of real world truth for smart contracts. Their goal like many others, is to introduce a service layer to the blockchain, to serve as a trusted resource repository for smart contract applications. These resource can take many forms, but in particular they provide smart contracts with a much more trusted digital footprint of reality than a centralised institution.
Whats interesting to follow with these projects, are the new architectural elements being invented to solve the underly question of, how do digital entities reliably reconcile themselves with real world states in a trustless manner. A lot of this involves computationally embedding political and economic principles which are effective in some contexts, but will completely fail in others, thus forcing developers and founders to innovate not just at the technology stack layer, but also at the humanities layer. Thus one needs to pay very close attention to this layer of any solution, and just like with smart contracts, we need to be cognisant of solutions being driven merely on the hype and glory emanating from the fundamental technologies used, as apposed to the exact individual impact the said solution will have if deployed in the real world today.
Another interesting approach is to embed digital systems in the real world aka the Internet of Things. By giving smart contracts the ability to read and write directly to the real world, full automation can be achieved in some applications. One key application could be in driverless cars, a smart contract can have direct access to a cars sensors and actuators and be able to fully control the execution of a delivery transaction. But then again, these embedded devices become the central point of failure, thus requiring the need to some extent, export the same consensus mechanism( proof of “x”) used on the blockchain into this IoT layer. This approach could lead to some very interesting hardware innovations.
In conclusion, a lot more attention needs to be placed on building a trustless layer which can facilitate the flow of information from the real world of atoms into the digital world of bits where smart contracts reside. The innovation and solutions around this problem might prove more ground breaking then smart contracts themselves.
