Decentralized Proof of Unique Individual
Consensus is a hard process because you never know who you can trust. In my recent articles I have been discussing the concept of basic income. One of the most basic requirements for any basic income system is reaching consensus on who is an unique individual. Identity theft would steal the birthright of some individuals while counterfeit identities (sybil attacks) would defraud the entire platform.
Solving the identity problem in a robust and decentralized way is the key to solving far more than basic income. It is the very foundation of reputation, account recovery, lost password recovery, and even democratic polling. This problem, properly solved, would enable a massive amount of innovation and economic development.
Centralized Validation
The most basic form of verifying unique individuals is what our governments do today. They collect birth certificates, death certificates, and photo IDs. These documents link every individual into the global family tree in a globally unique way. Some governments go a step further and gather DNA, fingerprint, and other biometric identifiers that help uniquely identify individuals.
This approach works well and is quite effective if you trust the centralized validator. We all know that governments can create any number of fake IDs for their agents. If each fake ID was allocated shares that only real live individuals are entitled then the government can secretly embezzle money from society.
Multi-Party Authoritative Validation
This approach creates multiple independent agencies each of which attempts to maintain a database of unique individuals. Think of this as every country in the world having a unique ID for every person in the world. An individual would be required to get “proof of uniqueness” from multiple countries before the global consensus would trust that an individual is truly unique. This model is far more reliable than the single-party model because it requires two entities to collude to manufacture fake IDs. The more parties that are involved in the validation, the more difficult it becomes to collude and the more challenging it becomes to defraud.
Under this approach the masses trust a handful of trusted parties to validate data. This minimize the cost per-individual of validating uniqueness. With 2 validators, each validator would be required to know the full population. With 3 validators each validator would only need to know 50% of the population (assuming even distribution of validation) and that the global consensus was OK with 2 of 3 approval. With 5 validators each individual would have to get validated 3 times and each validator would have to know 20% of the population on average.
As the number of validators grow the cost for one individual to validate themselves grows, but the less trust needs to be placed in any one validator.
Massively Distributed Validation
Under this model every individual validates the people they actually know in life. Think about your Facebook friends and family. The problem with this scale is that you can no longer trust the validators because there are simply too many of them. Sybil attacks will spawn thousands or millions of fake identities.
Hybrid Solution
There is no problem with having a few trusted entities that can certify data so long as the public is able to corroborate the results and keep the trusted entities in check. If a trusted entity is discovered to be forging identities then they can be removed from their position and the identities can be revoked.
Every user could publish a profile page that includes a photo and solicit endorsements from friends and family and link themselves into the family tree of all humanity. Two or more trusted entities would certify that the person claiming the account matches the photograph and/or other identifying information. This could be implemented with something as simple as netverify and/or competing services.
For just a few dollars per person we could get everyone verified and linked in a public social network of strong identity. A blockchain would take these inputs and use it to reach a consensus about which accounts belong to unique individuals and then use that information for many different applications, including a basic income.
Belonging to such a network would have huge value to service providers. Imagine if Steemit could simply use your validated ID to handle account signups and prevent abuse! This problem is widespread and thousands of internet companies are faced with the challenge of identifying unique people from thousands of fake identities.
Privacy
Not everyone wants to have their entire life linked through a public profile. It is entirely possible to generate an anonymous set of unique pseudonyms. Systems such as digital e-cash and/or platforms like zcash enable the identity verifiers to issue anonymous tokens to users who validate their identity. These same verifiers can then allow users to redeem their verified identity token for a signature on an anonymous ID. The verifier would be unable to link the token to the real life ID and yet could ensure they never certified more unique IDs than exist. While an individual validator could forge anonymous IDs, it wouldn’t be practical to get multiple validators to forge anonymous IDs.
Currently only governments have access to databases of identity information and through that database they have the power to erase you from the system. It is time we decentralize identity and opt for public databases of basic ID. From this we can implement transparent and fair systems of governance and money.
We have a real problem! On the one hand "reaching consensus on who is an unique individual" and on the other hand, individuals who do not want to reveal their identity. We want to remain incognito, anonymous but know everything about others. Pressure groups are demanding the right to be forgotten and the right to anonymity.
And what makes it a real problem is they both are right yet at the same time mutually exclusive.
It is very possible to have multiple identities, some pseudonymous, with only one getting connected to the formal system. Although it takes work/care, that is already how we do things today, for those who use pseudonyms to express their politics for example.
It was interesting to read. Well the problem is described. The material is presented correctly. Post deserves attention. I want to re-read to familiarize yourself with the individual parts of the post. It sets out clearly and intelligibly. The problem outlined in the post appears clearly to our consciousness. Thank you for having clearly shown that this is the case. We had to get acquainted with it and know more
well @dantheman - you are starting slowly to make more sense. And I agree that Proof Of Uniqueness (POU) for validation of identity is interesting.
The world is apparently full of fake identities from birth-certificates, drivers-license, passports and death-certificates - heck, people even buy their diplomas online instead of attending school.
If you are robbed for your passport when you are abroad, the only way you can get a new one is by contacting your closest consulate (for emergency passports valid for 1 roundtrip) or your closest embassy (to obtain a real passport from the government of your country official)
Just like a lot of people create anonymous accounts in the world of crypto, governments do the same for their agents as you said - so what is stopping someone from producing fake birth-certificates every year, registering them as citizens living abroad and bring the identities back in when there is a customer or an agent needing to request one?
And the world has a real problem with immigrants coming to Europe and USA burning their identities and in some cases their fingerprints - which means nobody knows what reputations they have from where they are from.
They could be hardcore serial killers posing as victims of terrorism and guilty to be sentenced to many lifetimes in jail in the countries they are from.
I agree we should decentralize identity and opt for basic ID in a public database - and it could be SteemID.
Sincerely, and in all friendliness
@fyrstikken & @fyrst-witness
Or your could engineer your system so that there is no economic advantage to multiple personhood by moving to a patronage model like we have here at steemit. Then you don't need ID at all.
Exactly. That's what I thought this whole system was about. But because of a few bumps in the road we're talking about identifying people and photos? What the heck. What is happening here?
Steemit is just one example .. I think this is a broader discussion.
Not apply to Steemit, but Steemit could us it to give out larger signup rewards.
We should be focusing on real issues with Steemit at the moment. Less than 24 hours ago, someone bought $5500 USD worth of Steem from an exchange and couldn't figure out how to move it to their account. I would call that a show stopper flaw.
See https://steemit.com/steem/@transisto/brought-a-lot-of-steem-but-can-t-figure-out-how-to-deposit-it-in-my-account-on-steemit-com
See https://github.com/steemit/steemit.com/issues/628
Let's re-focus and stop letting our heads wander. There is so much more we should be dealing with right now than considering identity verification.
That makes sense. Focus on what has more priority or even take advantage of current "problems"! Maybe we must follow the Microsoft’s paradigm! They had the "problem" of illegal Windows copy’s at the start, but guess what (!), they just leaved the "problem" unsolved until they got to the point of mass adoption! (because illegal copy's actually helped to distribute the Windows OS to the masses MUCH quicker,so it became the standard on top of other OS that where better at the time!) Maybe we must take the "Dutch approach" to pretend there is not a problem as long as the benefits are greater(?) and get focus on them until it is the right time(?)... Actually double accounts and bots make our platform more appealing (at least for now,the early days), because it simulates ACTIVITY !!! Everyone is more comfortable to participate on "active" platforms! for example on exchanges with high volumes(bots), on online poker rooms with much players (bots again, no wonder all new online poker rooms use bots).... So the point really is... "allow" the current problems to work in our favour and focus on the POSSIBILITIES that are more important and much easier to implement RIGHT NOW ...
easier sign ups,
more user friendly platform,
better design,
gamification,
picture uploads,
profile,
private messages etc...
Allowing people to follow tags the same way people follow users would turn the experience closer to something that was proven to work, Reddit.
I find most of the content on steemit.com to be sub-par, mixing all subjects together does not help at all.
You're putting out lots of good ideas lately Dan, thanks for sharing these.
De-centralized P.O.U.I tied into the steem web of trust social reputation system would be amazing :)
Have any of you read this gem?
Long read, but worth it.
https://medium.com/@ConsenSys/tell-me-who-you-are-258268bf3180#.jeaa3fstc
Can something like this be implemented here?
I love the idea of properly distributed validation. However the one problem that I personally see or have with collecting biometrics or DNA is it's incredibly intrusive.
Then comes the problem that regardless of how many anti collusion checks are in place. IF it's profitable people WILL find a way. You can almost be certain that if there is some way to exploit a system it will be exploited.
Turn the model around so that everyone benefits when people try to acquire more than they are given and give nothing for nothing.
Remember the way things were around here before the rep system?
People earned money for their work, and some built bots to do tasks like welcoming new users.
The only reason people were really upset about the bots, is that the bots allowed their owners to earn more money than they could because they had found some loophole that allowed them to have multiple accounts while most people couldn't. This allowed them to get around artificial limits like posting limits and enabled them to earn far, far more. I mean look at @wang he became a whale literally by saying hello and giving good advice to new users.
If you take away the artificially imposed limits in the system, the system will work to find it's own equilibrium.
Unless you keep your own data and have it signed by the verifier. It's invasive if you're forced to give away the data.
IMO any kind of documents cannot be used as proof of identity. Documents can be forged.
Some time ago, I encounter a project (I believe it was related to ethereum) which approach whole problem from a different angle: one identity cannot be in two places at the same time.
They focused on validation of presence at certain point of time of particular people. They build an app (or had a plan to do so), which helped organize thousands of simultaneous video-conferences between people divided into groups of five random people.
The task for people was simple: interact with all those people on your conference to make sure, that other 4-people are real, and for example they are not having another validation-conference at the same time.
Of course this is not a ideal approach, for example presence of all interested in validation people was required at the same time all over the world. The good parts is... that you didn't had to give up your privacy - you could use a wig or even had a mask.
If anyone know/remember this project, please link it.
Body double with mask would trick that system.
only if all those 4 people will accept his proof of being individual. In case of any doubts this confirmation should not be given.
But as I said... this is not an ideal solution, but I just wanted to mention about this in this conversation :)
some recent work on online pseudonym parties, the latest design has 1-on-1 pairs, more resistant against collusion attacks,
https://steemit.com/bitnation/@johan-nygren/on-the-game-theory-why-collusion-attacks-are-not-a-security-risk
Obviously I must be misreading this... In order to use steemit, you want a photo of me in the near future, and have other verified people verify it is me in the picture? Even if a 3rd party "Validator" is the one to verify me and my personal photo or info sits on their hard drive somewhere?
Everything gets hacked on the internet. Everything. My photo, if not uploaded, never can get hacked, but it was never uploaded in the first place. I like that privacy.
How does this help free speech if I have to stick my face next to something I say that might be against the policies of govt, etc?
How would this deter doxing people, facial recognition by big brother, etc?
This is really pedaling backwards here.
I'm going to consider you are writing this tongue-in-cheek Dan to prove a point, maybe to someone else, who chronically thinks this is a great idea. This is really taboo thinking. I don't care how decentralized identity verification is... it's still a way to track, monitor, and corral people, which can be abused by a state entity at some point. If there is a will, there is always a way.
Not really, with an identity verifier they don't necessarily need to save that data; just verify it and sign it.
I see the "Hybrid solution" as just a thought exercise to get us to think about solving the problem in one way. It does not have to be perfect now. If "Hybrid solution" gains ground then that would obviously need to be addressed.
I think most agree about this.. But remember the anonymous layer. Any person take a picture and know your name, that is not a big deal; it is info that is widely available. It is only if you can tie that picture to activity that the person wants to keep private. Cryptography can provide a layer of separation between the two.
I disagree. It has to be perfect now. If it isn't it will be abused. If it isn't abused, it will be used by those who we don't want to assign rights to use it. (ie: state entity or corporations).
You cannot launch a system, get adoption and fix it later. It needs to be fixed and perfect before launch..
That's why the internet and email and spam and DDoS are so uncontrollable. They said "it does not have to be perfect now". What a mess has been created as a result. Especially when you have hardware devices as part of the system. You cannot easily change every operating system, every computer, every router, every firmware, every phone when you change interopability standards very easy.
The rush to launch proof of uniqueness in our lifetimes will not be solved.
To be fair, those systems were all designed during a time when every node on the internet was already well known and sysadmins had a duty to monitor abuse and nip it in the bud right away.
See also BOFH
If we attached civil liabilities to companies and their executives who failed to address abuse issues those issues would go away pretty much overnight. Otherwise it's just a cost center.
You can get the answer from the same post ^^^ ... (no?)
None of these methods work. I can give horror stories for each one.
Currently, a school cannot even verify its students. Children slip through the cracks and illegal aliens are enrolled all the time. And this is a small bureaucracy. Where you should be able to get to a person that knows if some unknown child is in their class. But, it cannot verify that a child does not belong there. After a handful of teachers, it gets too complicated to make sure you ask all teachers if this kid belongs.
So, just considering a small set of our problem, we find that none of these ways actually works. Children that have moved, are still enrolled. The same child is enrolled somewhere else. Children that aren't supposed to be there are enrolled. Keeping more children on the books is a boon for finance.
And then there is the problem children. The ones that get probation all the time. Are they out or in? Are they missing because they are dead, or because they are in JV? These kids usually don't want to be there anyway, so it is not in their mood to help keeping their papers straight.
Add to that people being born and raised out in the Okees, people showing up and disappearing all the time. And what will you do with walk ins?