Hacker News Reader

Comment by scott_s | Hacker News Reader

A people's lack of identity in a nation is a civil problem. Technology does not solve civil problems. I don't see any indication in this article how blockchain-based identities will solve the civil problems faced by the Rohingya people; specifically, if the nation they are in does not want to grant them citizenship or basic rights, I don't see how asserting one's identity in the blockchain solves that.

The article states: "Noor’s goal is to give Rohingya the power to reclaim their identities with a resilient system that their host countries will recognise, allowing them access to social programmes, legal rights, education and healthcare." But that's begging the question. The problem is that the state won't recognize them. Why would a blockchain based solution change that?

View on HN

It's valuable to remember facts today, before they're forgotten, even if today's institutions don't attach the desired significance to them.

Blockchain is not required for that. You could just use a normal database.

Who decides if it's accurate? Granted, you might not need blockchain anyway. But I do believe some kind of cross signing setup at least is a better answer than just a normal database.

Who decides if the blockchain based solution is accurate?

An illustration: you do not trust Amazon because of a TLS certificate. That certificate means that you trust the entity on the other end is Amazon. But it doesn't make you trust them. You trust Amazon because of a bunch of soft reasons: you have ordered through them before, and it worked; they appear in the media a lot as a place people go to buy stuff online; a lot of your friends and colleagues buy things through them and are satisfied. All of these factors go into your soft judgement of why you're willing to give Amazon money.

(If you don't trust Amazon, substitute them for some online entity that you do trust. If there are none, please accept that some people do trust online entities enough to give them money and expect to receive goods or services in return.)

This is what I mean by technology does not solve civil problems: in this case, a specific technology solves the specific problem of establishing an online entity is who they say they are. But now you're back to the same problem that exists with any human interaction: do you actually trust the other side?

Back to the database: if we're going to trust some individual or institution to be the arbiter of accuracy, we're better off just using a normal database with whatever encryption we want. Because once your system has that basic level of trust, you don't need blockchain.

And back to my original point: this set of information (in whatever form) makes no difference if the people in power don't care. This information will not make the people in power care. The problem with any oppressed people is caused by those in power abusing that power. Those in power have power because they have might and the belief of the people that they will use that might to enforce their power. Blockchain doesn't change that.

I listened to some kind of "sales pitch" for that system. Their selling point was not to somehow force the state of Burma to recognise their citizens. It was to create a peoples registry of identities, without the aid of a nation state. Exactly how they collect these identities, I don't remember. I fell asleep.

This raises the "oracle problem" which is that blockchain-recorded assertions about things outside of the blockchain are not necessarily correct (unless a community has agreed to treat them as definitionally correct).

If you have a trusted oracle for people's identities, you might just be able to trust that oracle's assertions (maybe that's what you're referring to as "cross signing") without also recording them in a blockchain, unless you're concerned that the oracle will tell the truth initially and then lie later on.

I meant on a very low level - a central database can be manipulated to effectively rewrite history.

If you have a web of signatures, similar to PGP, Keybase and a few other systems, you can't do that.

The same goes for a blockchain.

What you put in there is susceptible to the oracle problem, but that is another level. Both these things need attention.

> a central database can be manipulated to effectively rewrite history

That's true, and that's an important issue. But I might still ask, for each application, is there someone who's trusted to introduce records into the system but not to update them later?

(We can definitely envision applications where the answer might be "yes" -- often related to concerns about censorship -- but it doesn't seem like this is the case for every application.)

> It's valuable to remember facts today

The last thing you want to do to an oppressed minority in an autocratic state is compile a public membership list. This article is someone tagging their pet fad to a humanitarian crisis.

I doubt an identity system designed today would have JPEGs and plain text in a great big XML file. More likely it'd contain digital signatures of hashed documents, including structures that could verify but not reconstruct biometrics. So the blockchain (or "list," if you prefer) would be able to answer yes/no questions, but that's it.

>Technology does not solve civil problems

How does the printing press making books cheaper not help end widespread illiteracy? Why does better engineering allowing construction of sewers not drastically improve civil life?

Heck something as simple as street lights must have done a lot to make people feel safer going out of their homes at night.

Yes, technology can change society and culture over time, which has an enormous impact on civil problems. But rarely does one single technological solution neatly fit into society and all of a sudden solve civil problems - because such problems are based on soft things like trust in others, trust in institutions, trust in government and confidence in basic human rights protection.

The people can though. Centralization via the state entity can allow for efficient decicions (like corporate hierarchies) as everyone knows which direction to look for common consensus.

With a blockchain approach, rather than separate entities clamoring to be the source of truth in lieue of an ineffective state everyone can instead look to the blockchain. (Which theoretically has no agenda, distributed control, etc...)

I think you are inverting causality. Myanmar is denying the Rohingya people their identity because it does not want them in their country. It does not deny them participation in their country because they can't verify their identity.

Let's assume this blockchain-based identity exists. What problem is now solved? What does it matter if people can now point to the blockchain to say "Yup that's me"? The state has a monopoly on force - that's rather what makes them "the state." Why would people look to the blockchain rather than the entity with a police force and military?

So it's more solving the wrong problem than solving the problem wrong.

Yes, completely. Which is why I personally find the proposals so baffling.

Reply on news.ycombinator.com