Proof of unique human


The way you use SAFE Network doesn’t have to be the way the other 7 billion people may use it.

The way forward for you is not necessarily the way forward for everyone.

A way forward against a corrupt authority is what we both want but your way forward isn’t necessarily the same way forward that I would take. I think there are many ways forward but that you will not get anywhere if you have limited flexibility. If SAFE Network can only be used for one niche purpose for a marginal demographic then good luck pushing your way forward by yourself.

You don’t represent the entire SAFE Network and you don’t represent the majority of human beings on planet earth.

The views differ on how much anonymity is necessary and for what reasons people desire to be anonymous. A purpose of having a human identity is so you can have elections (if you believe in democracy). Most human beings appreciate democracy and while I might agree with you on not wanting authoritarian government I’m not a person who thinks it’s possible to have a society without any form of government at all. If there is going to be some form of governance then you need reputation, voting mechanisms, and a way to prove you’re not a machine.

If you still want to fork it then go ahead and fork it. But I think it’s a waste of time to create two separate platforms just because you don’t want to use SAFE Network in a particular way. For example if I decide that I want anonymity with certain people but not with other people then it should be flexible enough that I can choose the level of anonymity for every interaction and if it’s not then I will use the fork which offers the most flexibility (for technical and statistical reasons).

Let users have choice and control over their identity and anonymity.

If you don’t want to be pseudo-anonymous then don’t identify yourself and let a random number act as a name. It’s not like you’re forced. On the other hand some decentralized applications require it and the vast majority of the people of earth actually want to be able to have some kind of identity. Why not let them do what they want?

Most human beings want a reputation of some sort. While they might not want to give up pseudo-anonymity there is no reason why they have to. You can have pseudo-anonymous reputation systems so the idea that you must either by fully anonymous at all times or fully public with your real name and address is ridiculousness. It’s also ridiculous to expect that the billions of people who could use SAFE Network are all going to want to be anonymous all the time.

The way not to have to deal with this pointless political bickering is to let the user decide for themselves how they want to interact on SAFE Network. If you don’t want to have a reputation or ID then don’t give yourself one but if you do then have one. There is no reason to have two separate networks because other people having a reputation or even putting their full name and address on SAFE Network should not affect you.


I think you’re confusing the network layer with the application layer. If I’m developing an app for human beings then I don’t want bots. I cannot get useful feedback from bots or AI.

I’m not against the idea of having AI do tasks but that is a completely different subject. The chief concern here is that for feedback whether it be democratic voting, or just a poll to see what new features human users want, then there must be a way to determine whether or not a participant is human.

If it’s a chess game as a decentralized app for example and you have people betting on who wins then of course you want to confirm that both players are human beings. If you don’t do this then you cannot have a chess tournament which pays in prizes.

This is the application layer not the network layer. The network layer I would agree with you has no reason to distinguish between human and machine. The application layer has every reason to distinguish between human and machine. Applications are for human beings while the network itself could be for the Internet of Things and not just include human beings.

So if your vending machine earns Safecoins then you don’t need to prove it’s a human. But there are situations where you want to know who owns the vending machine and a specific (unique) human being must be able to identify themselves to the network to prove they own the smart property.


Neither do you - I’m proposing something neutral, you are proposing that we are all id tagged to be users.

That would be great, but you are proposing that those that get tagged as ‘human’ have more authority and power over the network.

If individuals want to be open about there identity, then we don’t have to do anything - they’ll do just that. You are proposing a system wherein some users will have more rights than others and that is a recipe for authoritarianism.


What I propose is to screen out what is not providing value. Generally I think proof of work would most always screen out bots, but not A.I. – Why do you assume A.I. would offer no useful work e.g. have no useful feedback? Why is A.I. being discriminated against? I feel like there are deeply embedded assumptions here that you aren’t challenging.


I never said it applies to everyone. I said it’s a choice. If I want reputation and to be pseudo anonymous but you don’t then we should both get our wish.


Proof of Work doesn’t make since when SAFE Network has no blockchain.
I never said AI could offer no useful work but what kind of Proof of Work could an AI do? Image filtering? I don’t understand how you would do it technically.

I’m not entirely sure of what your argument is when you say AI is being discriminated against. Are you implying that it’s self aware?

I said this should take place at the application not the network layer. If you don’t install the application then you don’t get to vote at all.

I don’t know where you see authoritarianism. Are you saying any kind of voting leads to authoritarianism? That is a political debate which doesn’t belong in a technical discussion such as this. Being able to prove you are human doesn’t automatically produce network wide authoritarianism nor does it endanger network neutrality if it’s not a network layer function.

If it’s an application layer function then certain applications might become authoritarian or might be designed in an authoritarian way and then you just don’t have to run the app.


Not proof of unique human, but techniques for proof of unique computer. May be useful.


@HappyBeing This method may be better: remote physical device fingerprinting,

The real question is what purpose is there to identify a computer? The use case for identifying unique humans is so they can vote or so you can defeat man in the middle or similar attacks. Identifying unique computers has no legitimate purpose that I can see except to remove anonymity. It’s probably going to happen anyway but I see no reason to treat it as a feature.

The security risk is too high. Doing it would result in a unique computer identity database and then be traced to each person the moment they use the same computer to go on the regular Internet. It would remove all anonymity.

Having said all this it’s more than likely the people who use SAFE Network will believe it’s anonymous. It would only be anonymous if they can trust the decentralized apps as well as their hardware (neither of which is a guarantee).

Now if they are really clever they’d use a usb stick from someone else’s computer.



Suppose we can identify unique computers? What good is this? It’s good for tracking but SAFE Network by design is set up to make tracking difficult. If we could identify unique computers the only instance where it would make sense is for farmers. It might be used to prevent a Google attack or centralization in farming if the protocol could know how many unique computers are connected to the network.

But it would have to do so very carefully so that it is restricted to a true or false. Is the computer unique (true/false). Nothing more than that. No databases, no tracking, no storing any details which can identify the computers. All SAFE Network would need to know is if the device is unique.

Then it could report that there is x amount of unique nodes connected to SAFE Network. I think this is something which could easily be coded in if there were a need to have that information but again it’s a risk. Just having that kind of code in the source code at all is a security risk.

Biometrics is different. Biometrics would identify you but not the computer. It could be coded to confirm that you’re you so that no one can go and impersonate you. It would allow your friend to know they are talking to you and no one else. For certain kinds of applications this is essential and while it could reduce the anonymity of the user the user could remove the biometrics and their devices would remain unidentifable meaning they can go back to anonymity. If your device is identified and then you try to go back to anonymity it’s impossible. You go to Google and it matches your device to your identity on Google which alerts whomever has you under surveillance to check their databases of devices from SAFE Network.

There are ways around all of this. One way is to not use the same computer you use for Google for SAFE Network. Another way is so install SAFE Network on a raspberry pi type device. A final way is to design SAFE Network so that it can only identify your computer with your permission. It may be possible to use a sort of secure multiparty computation or advanced cryptography (zero knowledge proof) scheme to allow you to prove your computer is unique without revealing the details. (example: )

The truth is whatever you do on SAFE Network is going to leave a digital fingerprint. It’s unavoidable unless you take extra paranoid precautions.


@luckybit, I think @happybeing is just saying, Here’s a neat idea on how these guys solved a type of uniqueness problem. Maybe we can use some of this idea to solve the human problem.

Are you familiar with the “Yes, And” improv game? It’s extremely helpful in writing, I’d imagine it works for technical problem solving too.


The issues you raise are valid, but not applicable for the reason I posted, which I apologise for not explaining!

The value of proof of uniqueness (human or computer) is to prevent multiple account/node attacks. For example, solving the “Google Attack” posited in the Peter Todd thread, can be solved because proof of unique computer allows nodes to identify multiple nodes running on a system, and group them together.

At least in theory it might! I suspect that an attacker could falsify uniqueness quite easily.

As for tracking etc, I’m not clever enough to say wether that could be mitigated or not.


1 heartbeat 1 vote using something like the Nymi.

I was going to post this biometric idea too. It seems like a cheap, dependable technology that could be designed to offer a significant level of privacy if desired.

It seems to double to create an API with SAFE and have a pre-configured version similar to what LastPass and Yubikey do. I love my Yubikey, but I still have to always think about securing the device because anyone can press the button. Yes, it can be set up with a password (And secure stuff I do have them), but it defeats a large portion of the purpose for me. Having this 3 factor bio-metric device seems to be the next logical step.

I think this should be one of many different ways someone can prove they are human (And technically with this specific **biometric is could be fairly specific in proving age, since children have a very different electrophysiologic state that would very difficult to fake) and that the various approaches can correlate with the step-wise level of access/use of the network, which I am loving the idea of Browser, Farmer, Network Resource User…like things.


The main point for unique humanness could exist only to limit the over creating of accounts within too short of a time period. It is like a proof of work that only a person can do; therefore, limiting the number of accounts, the fatigue of a person over a certain number of accounts made. I can solve captchas all hour long, but I won’t solve mathematical problems for a second hour to make accounts. Especially if it is timed to 10 minute intervals between each mathematical problem… and I have solve three, that means only per half hour can I sit at the computer and perform account creation… just for example.

And Secondly, importance can rest on making a profile as a person within the SAFE network. So how to prevent another from taking my identity in the physical world within the SAFE Network digital world.


I commented on this in another Topic where I said I can create 000’s of tasks and dispatch them to micro-taskers who will each create one account for me. That would cost me let’s say 10 cents per task and just $100 for 10 TB of free data storage (if each new user gets 100GB free).

I have been arguing that instead of spending dev time on schemes that will be gamed because the incentive to harvest free GB’s in millions will always be larger than time that can be invested to prevent it, we’d be better off if we removed freebies (I said before, 10KB seems generous enough, but since that would take a whole 1MB block, maybe it’s pointless to give away less than 1MB per person) and instead invested time to motivate people to earn some storage space by doing something useful.


Reducing barrier of entry is quite important hence the free account and network access;
yet with microtasks which you mentioned if the tasks are sophisticated enough, and even disability mods are available and equal in time consumption, your India labourers for 10 cents per task will get quite expensive, and then what are you going to do with 000’s of accounts anyway… that’s where scenarios can get interesting.

It is clear that we can discuss issues regarding a single entity owning 000’s of accounts.
If I had 000’s of accounts what advantage do I have besides free space.?
Wouldn’t that be possible that 000’s of useless accounts can occupy too much of the network storage space, then bots can be designed to manage each of these 000’s of accounts to occupy bandwidth…
yet with a 30 minute delay it could take you months to get that together, plus sunken expense.
Unused accounts will increase the cost of network storage.


What other advantage should I need besides that? Amazon is $0.0300/GB per month and here there are giveaway ideas that go up to 100 GB of free space per soul ($0.10/forever). So over the first 12 months I spend $1 on my 1 TB of free MaidSafe space which I can front with my fancy service and resell at Amazon rates of $300+ year (it’s a bit more complicated but that’s the gist of it).

So I buy MaidSafe space from microtaskers for $1 / TB (forever) and re-sell it for $300 per year. Or $2, which is still handsome profit for my DropBox clone.

For the paying users and farmers they certainly would.


Interesting - and to think I give son#2 a hard time for spending too much time on Runescape - he has a very advanced character - and lots of homework to do…
I think its time to actually watch him in action in this game and pay a little more attention.


How about letting them watch advertisements to earn storage space?



I assume that was a joke for Warren!

I agree as long as I’m the one who gets revenue from those ads!


Sounds great. What stops the network from censoring people or flushing “undesirable” (obscene, politically incorrect, otherwise censored) data? Also while I know maidsafe allows for anonymous communicaton does it allow for anonymous searches? Call me paranoid but once you can verifiably tag something with a certainty then my next question is what stops some establishment wiping out all content according to that tag OR tracking the usage of content of that tag? Say we had a tag for “Ebook 1984” and some group wanted to get that off maidsafe or failing that track everyone that downloaded it or searched for it. Are users protected from that kind of invasion of their privacy and having their data and their searches compromised?

Doesn’t safecoin have a blockchain? Set up a small app to do the proof of work. Every proof of work task the user gets paid a small amount of safecoin. If you wanted negetive profit then if there were 10 tasks in a session then the user would pay 10 units of safecoin (10 safecoin, or 0.10, or 0.010 or whatever) and each task would be worth 1 unit, then when they were done they would have received their money back, similiar to how you insert a quarter to get a grocery cart and if you return the cart you get your quarter back. Or make it even simpler, pay 1 unit of safecoin for each question and get it back each time. I’m still a little fuzzy as to why the blockchain is needed but that would solve the issue I think.