4 ms·

As I understand it, You can't anonymize (remove) voter data and also validate that data. You're trusting the point of entry of the data where the ballot is subm

As I understand it, You can't anonymize (remove) voter data and also validate that data. You're trusting the point of entry of the data where the ballot is submitted. Simply all the blockchain becomes is a record that X number of people voted for Y. Whoever controls the point of entry controls what is entered into the anonymized data, and by definition cannot validate against it.

vmception 6y ago
> You can't anonymize (remove) voter data and also validate that data. There are a variety of teams and projects working on that problem. I wouldn't say that limitation is really an absolute.
- im3w1l 6y ago
  I think under fairly general assumptions, a purely digital voting system cannot work. But it may be possible to bypass those issues with a hybrid physical/digital system. Or it may be possible to construct a system that is unsound in theory but works decent in practice. I'm thinking about technologies like DRM and Proctoring. But deep fakery could become a threat to proctoring in the near future.
dcolkitt 6y ago
> You can't anonymize (remove) voter data and also validate that data. Sure you can. That's what the cryptographic field of zero-knowledge proofs is about. For example there's well developed protocols where every voter can verify that their vote was counted exactly once, anyone can verify that no non-authorized voters were counted, and perfect secrecy is preserved.
- hatsunearu 6y ago
  You'd need homomorphic encryption to achieve this, not ZK.
  - sebmellen 6y ago
    Not if the vote counter is party to the ZKP as a verifier, but then we end up with a centralized source of trust anyway.
  - catlifeonmars 6y ago
    Homomorphic encryption could be seen as one method of implementing a zero knowledge proof. Traditionally, zero knowledge proofs are implemented with more statistical methods however. As an aside, having the verifier choose the desired confidence interval is a rather nice property of interactive zero knowledge proofs.
fabianhjr 6y ago
You can anonymize ballot data and validate it via ring signatures. (At least it is one proposed way of doing it) It has the drawback that each person needs to know a bit more about cryptography than a layperson should so it is something that is currently not accesible for general elections.
- bigbubba 6y ago
  As I see it, a key element of democracy is making people confident that their opinion was counted. We ask the masses for their input not because we think mobs of people are particularly wise, but rather because the populace being consulted, and truly believing their votes were counted, facilities reliably peaceful transitions of power. The social stability provided by elections perceived as legitimate is good for economic prosperity. A legitimate election that 99% of the population doesn't understand or trust loses this utility despite still being legitimate. An illegitimate election which people mistakenly believe to be legitimate is actually better than a legitimate election which people don't understand or believe.
  - fabianhjr 6y ago
    Yeah, for example with First Past The Post people have an understanding that that is "how elections work" even if simpler alternatives are more expressive. (Eg, Approval Voting/Combined Approval Voting) There are some countries that deprecated their FPTP method to great results.
hutzlibu 6y ago
"You can't anonymize (remove) voter data and also validate that data" It probably is a very unpopular opinion, but what about thinking again about an open vote? What is wrong with openly standing by your choices? The fear, that there are too many bullies around who would bully people to change their vote? Well, if bullies do this, they also do other things, so we need to protect the vulnerable of them anyway. The way I understand it, the secret vote comes from a time where henchmen of the local warlord/prince/ were much more common and secret vote is a defence against organized bullying to votechange. But are we( as in people of western states) still living in such times?
- arrosenberg 6y ago
  > What is wrong with openly standing by your choices? People with money will buy votes if you make it possible for them to know how people voted. We have historical examples in US elections pre-1885 and in Congress post-Watergate.
  - hutzlibu 6y ago
    Well, I would let them sell their votes. If enough people of a society do not care about their vote, so it has a significant impact, than it is probably not a very stable society in the first place.
    - arrosenberg 6y ago
      Probably a good thing you don't get to make policy then. Not everyone is able to get enough of an education to appreciate what you have. Good democratic leaders make rules that account for that and preserve the vote.
      - hutzlibu 6y ago
        "Good democratic leaders make rules that account for that and preserve the vote" So good democratic leaders make rules to protect the people from themself? Well, I know that one. I simply disagree. I rather think powerful people make rules to let them stay in power. An obscure vote system can help with that. Imagine suddenly a change to open-vote: would Trump really have won, if he just would have tried to buy votes? I doubt that.

vmception 6y ago

> You can't anonymize (remove) voter data and also validate that data. There are a variety of teams and projects working on that problem. I wouldn't say that limitation is really an absolute.

im3w1l 6y ago
I think under fairly general assumptions, a purely digital voting system cannot work. But it may be possible to bypass those issues with a hybrid physical/digital system. Or it may be possible to construct a system that is unsound in theory but works decent in practice. I'm thinking about technologies like DRM and Proctoring. But deep fakery could become a threat to proctoring in the near future.

dcolkitt 6y ago

> You can't anonymize (remove) voter data and also validate that data. Sure you can. That's what the cryptographic field of zero-knowledge proofs is about. For example there's well developed protocols where every voter can verify that their vote was counted exactly once, anyone can verify that no non-authorized voters were counted, and perfect secrecy is preserved.

hatsunearu 6y ago
You'd need homomorphic encryption to achieve this, not ZK.
- sebmellen 6y ago
  Not if the vote counter is party to the ZKP as a verifier, but then we end up with a centralized source of trust anyway.
- catlifeonmars 6y ago
  Homomorphic encryption could be seen as one method of implementing a zero knowledge proof. Traditionally, zero knowledge proofs are implemented with more statistical methods however. As an aside, having the verifier choose the desired confidence interval is a rather nice property of interactive zero knowledge proofs.

fabianhjr 6y ago

You can anonymize ballot data and validate it via ring signatures. (At least it is one proposed way of doing it) It has the drawback that each person needs to know a bit more about cryptography than a layperson should so it is something that is currently not accesible for general elections.

bigbubba 6y ago
As I see it, a key element of democracy is making people confident that their opinion was counted. We ask the masses for their input not because we think mobs of people are particularly wise, but rather because the populace being consulted, and truly believing their votes were counted, facilities reliably peaceful transitions of power. The social stability provided by elections perceived as legitimate is good for economic prosperity. A legitimate election that 99% of the population doesn't understand or trust loses this utility despite still being legitimate. An illegitimate election which people mistakenly believe to be legitimate is actually better than a legitimate election which people don't understand or believe.
- fabianhjr 6y ago
  Yeah, for example with First Past The Post people have an understanding that that is "how elections work" even if simpler alternatives are more expressive. (Eg, Approval Voting/Combined Approval Voting) There are some countries that deprecated their FPTP method to great results.

hutzlibu 6y ago

"You can't anonymize (remove) voter data and also validate that data" It probably is a very unpopular opinion, but what about thinking again about an open vote? What is wrong with openly standing by your choices? The fear, that there are too many bullies around who would bully people to change their vote? Well, if bullies do this, they also do other things, so we need to protect the vulnerable of them anyway. The way I understand it, the secret vote comes from a time where henchmen of the local warlord/prince/ were much more common and secret vote is a defence against organized bullying to votechange. But are we( as in people of western states) still living in such times?

arrosenberg 6y ago
> What is wrong with openly standing by your choices? People with money will buy votes if you make it possible for them to know how people voted. We have historical examples in US elections pre-1885 and in Congress post-Watergate.
- hutzlibu 6y ago
  Well, I would let them sell their votes. If enough people of a society do not care about their vote, so it has a significant impact, than it is probably not a very stable society in the first place.
  - arrosenberg 6y ago
    Probably a good thing you don't get to make policy then. Not everyone is able to get enough of an education to appreciate what you have. Good democratic leaders make rules that account for that and preserve the vote.
    - hutzlibu 6y ago
      "Good democratic leaders make rules that account for that and preserve the vote" So good democratic leaders make rules to protect the people from themself? Well, I know that one. I simply disagree. I rather think powerful people make rules to let them stay in power. An obscure vote system can help with that. Imagine suddenly a change to open-vote: would Trump really have won, if he just would have tried to buy votes? I doubt that.