Proof of unique human


Like @Andrew_Masters wrote, the POIs are only valid during a certain time-frame. Each POI is valid until a next set of POIs have been issued one month later, and when you get a new POI, that new one is not traceable to your previous one.

The protocol is similar to what Bryan Ford described in his 2008 whitepaper,

​_We propose Pseudonym parties, a scheme for creating accountable pseudonyms, which combine inperson social occasions (parties) with technical infrastructure (a pseudonymous sign-on service) to enforce the rule that one real person gets one virtual persona on any participating online service. Pseudonym parties enable the user to adopt different personas in different online spaces without revealing the connection between them, while ensuring that each user has only one accountable pseudonym in each space. Pseudonym parties can be started incrementally in a fully decentralized fashion, can run on volunteer labor with minimal funds, and may even be fun._​

The POI tokens are not by themselves tied to an image of you, or a name, or any other identity-data. Since the tokens are not linked to ‘who you are’, but rather to the proof that it’s very hard to obtain multiple of these tokens, they could be understood as being anonymous tokens. These tokens can then be used to sign identity data like a uPort ID or other IDs, or used to gain access to governance services such as crypto basic income.


The unique human concept is a dead-end if the purpose is to keep an infinity of bots out. The remedy to that threat is economic, such as payment, or POS or POW. Then you will have a few (or more than a few) bots amongst the humans, both of whom will be able to engage in smart contracts.


The POI protocol uses deposits to remedy the risk of being attacked by an infinite number of bots.



Could you expand a bit on how the anti-spam mechanism works?

The randomization makes it hard to gain control of entire groups, but controlling 2x or more the total number of users is a way to bypass that. Anti-spam deposits make it very expensive to perform this attack.


Yeah, if making an account that can do important stuff costs an investment of time, money and/or effort it strongly incentivises against multiple accounts. The issue is the on-boarding process can’t be too hard or people won’t join to start with.


Thanks (I really want this to be solved, it would be superb) for the info. I am still not “getting it”, please don’t think it’s a critique though, I am always on the other end of this conversation :slight_smile: So the POI is a group and that group are all identified as human but not necessarily unique. Is this correct?

Then the group is re-created every X days (say month).

So then if the network is millions, what is the process (prob invalid question if not trying to be unique)

Thanks again for input, not sure if you are involved, but appreciate the info regardless, it’s an important issue and a really interesting one.


Could only post 1 image and two links, so put the post on a gist


you can use a hologram :slight_smile:

The protocol could easily be adapted to let people use holoportation for the hangouts, as an option besides with the current webRTC hangouts. It would be an improvement, and allow for more immersive co-presence.

FAQ: How do pseudonym parties work ?

The proof is based on the fact that people cannot be in two places at once. It was first documented in 2008. The POI protocol does meetups in the digital space using video-hangouts, and could be modified to include holoportation.

"pseudonym parties currently appear to be the only proposed solution that can address both strong accountability and privacy at the same time. "

-Bryan Ford, 2007, Pseudonym Parties: An Offline Foundation for Online Accountability



No, it is a ridiculous idea. No-one except a fanboy clique will use the SAFE network if such a requirement is implemented.


I would highly doubt anyone would put such a requirement on creating an account with safe. I expect multiple accounts will be a feature of the safe network. I could see some voting apps using something along these lines though. Especially if each video chat was totally secure in vaults and not avail to anyone except those in the chat (and the recognition software in my model’s idea)


New article about the project :slight_smile:


Wrote a short whitepaper for the Proof-of-Individuality (POI) system


So what stops me from doing one “party” at 2pm and another at 5pm and becoming 2 identities in this system. Any ID (licences etc) can be faked. So then I become 2 people.

The premise that you cannot exist in two different spaces at the same time is invalid.


all hangouts happen at the same time


And why can I not be in 2 different spaces at the same time?

AI is becoming better all the time and with morphing software I can appear as multiple people at the same time.

This certainly has a limited time before its defeatable and even with real time morphing software it can be attempted now.


ritual memes will evolve, and people will learn to recognize those who join multiple hangouts. if you spread your mind over two hangouts then you’ll most probably fail to appear as you would appear if you gave one hangout your full attention.

AI that can pass as human won’t exist until after the techological singularity.


That is true and the idea directly threatens privacy, one of main goals of the system.

I can’t believe this discussion is still going on.
In theory, PoUH shouldn’t be discussed if the system was well-designed.
We hear PoW is wasteful and inefficient, and yet this topic keeps going on.


Ah, but human assisted AI exists, real time morphing and very simple 80’s style human assisted AI (or AI assisted human if you wish) does exist and could be used.

My point is that it can be fooled and relies the 4 others to spot it. But if you do it world wide at teh one time you get some very sleepy people and cannot expect great responses from them. So no I would not trust this method to be good enough for the intended purpose.


sure you can morph your avatar, but you could also just appear in your own body for a few minutes and enjoy what the system provides, decentralized sybil proof using strangers…


And consider the opposite side of the coin to the problem of human-assisted AI: Some genuine humans, I’m thinking of Aspergerists so prevalent in tech, have trouble appearing as fully human. The people in their group would be like “hey, are you really human?” and want a closer and more extended look, which is very distressing for most people.

This whole discussion is an infinite regress. Just accept bots but make them do something (POW or whatever) to eliminate an infinitude of them.


you’ll be interacting with strangers, over an encrypted transmission, the token you receive will be anonymous. each new token you receive is not traceable to your previous one. the anonymity behind it is similar to Bitcoin as ‘digital gold’, the tokens are public, but it’s almost impossible to trace a POI to whoever holds the private key, and each new token is registered on a new private key.