Serious uses for blockchains with smart contracts?

Question

It seems like many of the canonical "use cases" for smart contracts don't hold much water under examination. Typically the issue stems from the fact that the points of contact between the real world and the digital state need to be centrally managed or cannot be trusted.

The assumption that if it depends on a central authority it's a waste of time is made throughout.

I'm not asserting that all use cases are never going to work, I'm just looking to the community for feedback on the following doubts about commonly cited applications:

Applications

1. Digital Identification / Voting
2. Property / Land Ownership
3. Medical

Problems

1. How do you assign each and every citizen ONE keypair without a central authority?

2. I don't believe people would seriously wish to live in a world where rights to land and property are stored immutably on a public ledger. This means if guys in balaclavas force you to transfer the farm under duress, there is no central authority to reverse this.

Smart contracts can aid the medical industry in terms of managing and sharing sensitive patient data. This can help when conducting clinical trials, by automating and tracking patient consent, ensuring privacy of patient data, streamlining setup processes, and facilitating cross-institutional sharing of information. Similarly, the same benefits can be used for the conduct and sharing of life-saving medical research, such as cancer research. By addressing privacy concerns and increasing transparency, institutions might be more willing to collaborate and share data with one another, helping to boost important research in the field.

In a similar vein to 1) how do you know the blockchain address corresponds to the patient standing in the consulting room unless keypairs are given out against a passport/driving license by a central authority? Without this, the data is of unknown integrity. If network participants may generate addresses autonomously and start appending arbitrary medical histories, why should anyone trust the findings of analyses on the data?

Answer 1

Many governments around the world have started working on the concept of "Digital Identity".

In a nutshell, the government would assign you a private key at birth instead of or in addition to a social identity number or whatever they use to identify you now. This private key would allow you to prove your identity to any third-party (private or public), store encrypted information about yourself and give consent to share that information to others. You could think of use cases such as sharing medical record, proving your identity at a bank and accessing your bank account online, signing a contract, storing and sharing your school grades, etc.

The issue that you raised are valid: The private key is initially assigned by a central authority, the government. This is basically the oracle problem, where someone must be trusted to transfer an information from the outside of the blockchain to the inside of the blockchain (in that case, the outside information is the link between the natural person and the public address derived from the private key).

The government should not know your private key. It should only know that the public address derived from the private key belongs to which natural person (or maybe there is a way that it does not even know that). The natural person should be the sole owner of the private key and of the information accessible by that key.

Another issue is that the private key should not be a password that could be lost or stolen. It should be something that only the person has access to, perhaps something biometric.

A third issue, like you mention, is how to reverse transactions that were made under duress (your example of transferring property when being threatened)? You could also think of exercising seizure when a court orders it without the consent of the owner, which is also important. There are solutions to this, such as granting judges the authority to change property of an asset. The judge does not need anyone's private key - the property smart contract must only allow a judge to change property without the consent of the current owner. Of course, you need a central authority to assign power to a judge (the government), but this is already how our legal systems work.

I think that we are at least 15 years too early for any implementation of something like that. There are of course many, many issues that need to be figured out.

Answer 2

after weeding out the questions from your question, I see the following:

How do you assign each and every citizen ONE keypair without a central authority?

Well there is no requirement as such to have only ONE keypair. do you have only ONE bank account? do you have only ONE social media profile? these are centrally operated industries and yet they can't enforce such a one-per-person policy. even large scale government databases are famously known to issue multiple id's to the same person

In relation to voting, this gets even more interesting. There is a reason with all the technical advances of the 21st century, I can order a taxi at will, however I still cannot vote electronically. Other than the obvious political issues, voting even through the internet is a hard problem. one which will not only be solved by blockchains alone.

I don't believe people would seriously wish to live in a world where rights to land and property are stored immutably on a public ledger. This means if guys in balaclavas force you to transfer the farm under duress, there is no central authority to reverse this. (I'm presuming theres a question in there somewhere)

your assumption that the arbitrager in such scenarios has to be centrally operated is untrue. for instance the case of chainlink, bringing off-chain real world data to the blockchain is a hard problem, and the entire protocol is designed to solve that without the need for a centralized third party.

In a similar vein to 1) how do you know the blockchain address corresponds to the patient standing in the consulting room unless keypairs are given out against a passport/driving license by a central authority? Without this, the data is of unknown integrity. If network participants may generate addresses autonomously and start appending arbitrary medical histories, why should anyone trust the findings of analyses on the data?

google "Non-interactive zero-knowledge proof's" for a possible solution to that