Proof of unique human


Yes but how would one code one’s unique subjective emphemeral entropic experience in time’s flow? And to think we’re discussing something that sounds like it belongs in a science fiction novel just to stop bloody spambots from taking over the net.


Robots could soon read your mind - and there are no laws against it

Nita Farahany, professor of Law and philosophy at Duke University, said: "Beyond two-factor idenitification…is using neural signatures - ‘pass thoughts’.

"It turns out every person thinks quite differently about the same thing.

"So you could think like a song or a little ditty in your head while you are wearing a consumer based EEG device and then that, which has a unique neural signature, can be used as your pass code.

“It turns out that’s an incredibly effective, incredibly safe and almost impossible to replicate pass code, so there is discussion about using pass thoughts.”


if any parsing or input can be faked and if cumputers will advance as fast as we think they will then really one has to ask is there a point to trying to create a pouh? I mean given time an A.I. will be indistingishable from a human on the net. And if we throw things like neural interfaces into the mix that can read our minds one really has to take security breaches seriously. It’s one thing to get your password it’s another to get inside your head. Granted the uniqueness of thought could be very useful as a security measure in itself however if you literally store your thoughts as data then that could backfire quite severely.


A low tech approach to Proof of Unique Household might be to create a service that requires you to upload an image of your electricity meter and the name of the website you are registering with (leaving aside those without electricity meters for now), in exchange it gives you a token to prove this is a valid recent electricity meter image.

The photo of the electricity meter could be compared against previous photos of allegedly the same meter. The meter has a unique serial number and lots of other distinguishing marks, and has the nice property that the number of units is always increasing, so that we know if one picture is older than another.
You then use this token when registering to prove you are a unique household.

The service would have access to the list of national electricity meter ID’s and recent readings, and would be able to tell if the image was an actual recent picture of the meter or a fake. Because the number of electricity units is always increasing, the same image could not be reused.

An attacker would have to either physically visit your house and snap a picture (unlikely but still possible) or somehow fake a picture.

Still some rough edges, but would welcome thoughts on this as it doesn’t sound too complex.

Also, as an aside, I don’t like the term ‘Proof’. Proof works well for maths, but for real world things we can only have ‘Evidence’. We need varying levels of evidence (from different sources), according to our use-case.


Two major problems with this.

The main one is loss of anonymity, you have to rely on an authorising agency.

The second and very important is a lot of people are going off-grid with solar and other methods.

And related to the second is that there are quite a number of houses in Australia who do not get electricity from any authority. Some of the properties these houses are on are extremely large, the size of some small countries. EDIT: I forgot uni colleges, some appartment blocks, many flats do not have separate metering.

And thirdly, I could go up and down the street at noon and take pictures of a few hundred house’s meters and there is no way to detect that. I would expect anybody wishing to appear as hundreds of households to do that. Not to mention that modern meters use LCD displays that from a photograph are virtually identical and no software can distinguish the difference from a made up meter (with random fake marks inserted) and a suitable meter reading.

My previous house had 4 meters, as do a lot, so I could pretend to be 4 and the rest have 2 meters.

So the uniqueness is gone.



##Proof of Individuality

POIs (proof-of-individuality) are smart-assets that are hosted on the Ethereum network. They solve one of the hard problems in crypto - how do you prove that a person only has one account within the system?

Join the POI project on Slack,

Through person-to-person verification. Users are grouped together by random in groups of 5 or so, and every group does a video hangout at the exact same time, that lasts around 10 minutes or so. Users check so that the others in their group aren’t doing another hangout at the same time. They then sign each other’s POIs and verify them. Once the hangouts are finished and all POIs have been verified, everyone will know that each POI represents a unique human being.


What stops you joining two groups as two identities at different times though? This looks like accounts are individual which would be a much easier task? Or am I missing something here?


There is one thing that will possibly give us proof of unique human. That is higher AI looking at a human with a broad spectrum sensor. I don’t think we get higher AI free of will. If AI is cooperative it may render human society completely transparent.


To me this sounds like proof of human. Not proof of unique human which is the real kicker. Yes, its nice to know that every person that has a PoI IS indeed an actual person. However, someone can have many different PoI’s and that breaks things like one person one vote (haven’t followed this topic for a while, but that’s what started the discussion if I remember correctly)


If there was advanced facial (and perhaps voice) recognition software running at the same time to verify the same person was not in multiple chats it could potentially work. All chats would have to happen simultaneously (I guess the chats could be recorded and verified against as well.) The five random users would just be verifying that the other four were really people. Similar chats (determined by advanced recognition software) could be marked for review by random members as well. Similar chat members could be required to be in the same chat at the same exact time for verification.

The more I think about this, the cooler what I’m explaining seems to me. Would it be 100%? It might be close to that if the recognition software was “smart” enough, and the community had a vested interest to see that all accounts were unique.


I like that, I like that!


I think you PoI is only valid for a certain timeframe, you have to complete another (completely separate) PoI once this expires (say every month or every year). Everyone has to complete the PoI at the same time.


If this works, it means we don’t have to waste stupid resources farming with GPUs etc. We can just say these are the transactions and digitally sign that I, a unique human say they are okay.

If you are a unique human, it makes it something really cool possible - universal basic income. As a ‘reward’, for being human UBIcoin could issue you with 100 credits each week.


Unless of course you generate your own power in excess of your usage and therefore your meter would run backwards thus disproving your theory.


Sounds laggy. It would also require each participant to have a webcam and enough bandwidth to pull off such a video conference.


Something very similar to this was discusses before on here :slight_smile:

I was thinking about another implementation last night (by chance) which is perhaps even simpler. It’s not fully fleshed out though so here’s just a very rough outline:

There is a token (we’ll call it PUH coin) which has some inherent value (for example it would enable access/give money off certain of apps which benefit from PUH).

I start with 5 PUH coins.

I then nominate 5 people/accounts which I believe to be human and my token is stored as a vote for this person.

That person in turn gets to place a further 5 votes on accounts which they believe are backed by a human, and so on.

As soon as you have voted on someone you start to receive PUH coins as dividend payments at a rate of 0.1 coins per week.

If the user you voted for has another vote of confidence from another user, your dividend payment rises, to say 0.2%. For 3 votes it would be 0.3% until a maximum of 5 votes from other users, and 0.5% dividend payment from that coin each week.

All arbitrary numbers, probably lots of problems/hole/extreme use cases, and surely something similar done before but still maybe a little bit interesting.


And what is to stop real time video morphing with voice morphing so that I can talk as two different people in the one video conference. Just not talk over myself.

Or talk to two different groups at the same time as apparently two different people.

The software is often used in animation so the operator can act normally and the character mimics facial expressions and movements. But for human morphing it improving all the time and on a lower res link it is hard to distinguish

Whenever you digitise biometrics the potential to mimic/copy/morph is always present.


Normally in when you are generating you have 2 meters, one for usage and one for generation that both always go up. Although this might not be the case everywhere.


Indeed no it’s not true everywhere which is kind of my point.


And meters can be replaced. The person moves house too.

Over the population it would represent a lot of situations. Not particularly suited as a proof of unique human, But of course shared housing, mobile homes, living in RVs all suffer from never having that one-to-one relationship between meters and person.