It should be possible to use plain old cryptography to prove
1. I am one of the named, publicly accountable people registered as participating in this thing, and the same as posted under this pseudonym yesterday
2. Provided I'm reasonably careful, you can't tell which one is me unless n of m participants agree to unmask me.
3. I can only post under one name at a time. I can change pseudonym, but then my old one is marked as abandoned, so I can't trivially fake conversations with myself.
Doesn't that then require a centralised ( or a hierarchy of centralized ) authority to manage point 3?
Who would that be? ( each country issuing it's own citizens IDs? )
If the solution requires you to keep a private key private ( to prove who you are ) - how is the average person going to do that?
How are you going to build in your cryptography into all those different systems?
All you need is one link between that pseudonym and some identifying info - like IP address or a payment - and it's all gone, and you've already built a perfect system for government tracking.
So even if you have built all that successfully I'd still suggest the world would split into sites that would use it and sites that wouldn't.
> Doesn't that then require a centralised ( or a hierarchy of centralized ) authority to manage point 3?
I'm not sure. I think it might be possible. Lots of wild things are possible with cryptographic primitives.
Even if a centralized authority is necessary for coordination (which seems inevitable anyway, the forum has to be hosted somewhere), if the authority could blind itself about who was who, that would be valuable.
Yes, any system which gives users the option to hard-ID themselves - which I think is a desirable feature, and probably necessary to prevent spam/sabotage - needs to piggyback on something existing, like a government ID scheme of some sort.
> If the solution requires you to keep a private key private ( to prove who you are ) - how is the average person going to do that?
Same way they keep their cryptographic keys secret today, their software does it for them with very little conscious effort. And yes, people can get hacked etc. I'm not proposing something that would solve all problems once and for all, just something that would be useful.
> All you need is one link between that pseudonym and some identifying info - like IP address or a payment - and it's all gone,
Yes, people would have to still be careful to not ID themselves accidentally. But worst case, we're back to more or less the kind of forum we're in right now with respect to identifiability (and probably still better when it comes to astroturfing).
> So even if you have built all that successfully I'd still suggest the world would split into sites that would use it and sites that wouldn't.
That's OK. I imagine lots of different "forums" or "chatrooms" with the feature that you know the list of participants, but you don't know which nym is which particpant.
So in summary:
- the requirement that one person only has one of this sort of ID requires a central issuing authority which can ensure this.
- in the end such an ID would still require very careful use to remain anonymous, though I do like your idea of being able to change it in a one in/one out way.
Isn't it much simplier to have sites which link to real ID's and sites which implement their own throwaway pseudo-anonymity - and all we need is it being clear when real ID's are being used?
Is there really any benefit to conflating the two?
Yes, if we can manage it. The benefit is that we could talk about controversial topics, e.g. foreign wars, without worrying that it's 20 fake people trying to convince the one real person - while giving the powerful minority as little information as possible about who's criticizing them or organizing against them.
> while giving the powerful minority as little information as possible about who's criticizing them or organizing against them.
I find it hard to believe that a system that requires a central issuing authority in order to work does not simultaneously create a perfect tracking mechanism by those in power.
Also, unless you take extreme measures, I think it's best to assume that if the government wanted to look pretty much everything you say on the internet is tracable to you - you can't escape the physical aspects of the internet - in the end, for people who have access to the hardware, I find it hard to see how you can simultaneously remain anonymous and have your traffic routed.
In the end if you want accountability for content that other people say, then you also have to accept that accountability yourself - real ids.
For example, let's say your scheme worked - I could hire out my identity to a lobbying company - not a problem as the content won't be linked to me - but I get paid while there bot posts stuff on my behalf. At the end of the campaign I change my nym and break the connection between the last campaign and the next.
However if that was a realID all you could do is be a paid lobbyist - but you'd have to be personally comfortable with what you are saying and you'd find it harder to pretend to be a veteran speaking from experience or whatever.
It's also the flip side of people feeling free to say what they want under the cover of (pseudo) anonymity.
I wonder if one solution is to partition the web into places where anonymity isn't possible, and places where it is.