Bitcoin as a viable model for Anonymous, valuable identities on the internet?

I hate spam. Anonymous people wasting your time and resources for their commercial gain with no way to track them down and blacklist them is a significant problem. Any identity of theirs that you blacklist is so cheap that it is instantly replaced with 10 new ones.

I also hate all current services that establish identities but do so in very dangerous ways. Users are asked to invest time creating these identities by websites, ostensibly either for the purpose of building community context about each other (for instance Facebook profiles), or as a way of preventing spam (Disqus logins). Many sites are simply looking to build human or human looking identities in order to sell them to advertizers. These sold and resold private identities, or the more limited yet more sinister dossiers compiled on us by various law enforcement agencies, are becoming dangerously complete weapons aimed at any concept of a private life for people on this planet. When Eric Schmidt said that privacy is dead, he underscored just how far business models that sell our identities as goods will go to turn every atom of who we are into pennies bouncing through a global Rube Goldberg machine.

So, what we need is a concept of identity that will subvert both of these problems. Have we already seen the solution, and just not recognized it?

Lets use the mathematical ideas of bitcoin to solve the internet identity problem. We want identity to be anonymous where needed, decentralized, with no central authority allowed to be the final arbiter, and no advertiser able to turn our internet name into a product to be sold. At the same time, the problem of SPAM, and the need for some trusted identity for community building teaches us that identities should not be so cheap and disposible that one can use them for antisocial or fraudulent purposes without reckoning some cost.

In this system, (call it Bitvoice?), identities would be created based on solving some number problem similar to the hash collisions used by Bitcoin. This would have to be a significant effort, perhaps somewhat easier that discovering a bitcoin, but significantly more costly than just generating an RSA public private key pair. An algorythm should be selected that would generously outpace, but not totally become unhinged from expected population growth of humans (and any other intelligent beings with identities we expect to need to communicate with). The resultant solution would represent an identity that could be used to communicate online. The identity should be cryptographically identifiable as the same identity, should be not inherantly traceable to a human except through the content of communication it chooses to produce, yet should be expensive enough that the blacklisting of this identity as a spammer or a troll would be a small, but significant financial setback.

Dissidents or people who need anonymity could use newly computed or donated identities to do their work, while spammers, who require millions of fraudulent identites to eek out a fraction of a cent in returns would be quickly bankrupted or have to rely on botnets to slowly regenerate blacklisted identities. So, itentities could be tuned to the right scarcity for nurturing a well balanced online society.

The initial distribution of identities would be the big problem with this system. How can the average John Q Public, or the average child in Mumbai get a small number of identities to work with initially, with some being able to be registered for government services and benefits an others able to be kept anonymous?

Anyway, I think there might be something to think about here. What if we could actually find a technical framework that would balance anonymity and identity towards a happy medium? That would be a good day indeed.