15 comments

[ 2.9 ms ] story [ 40.0 ms ] thread
I'm glad someone else found my paper interesting enough to post here. I'd love to get your thoughts on the idea.
It seems pretty solid but I'm skeptical that you could ever get regular ppl to be able to use and understand a crypto-system like this.
Aside from obvious things like the immaturity of the various technical components (blockchains, biometric security), I think you’ll have a hard time convincing folks outside small circles of crypto idealists that sampling democracy is desirable. Even if it works at a technical level, most people want to explicitly chose who reps them and makes laws for them. Randomly selecting a sample to vote on a law takes that away, and isn’t a direct digital version of contemporary representative democracies/republics. To channel Neo, the problem is choice.
I would prefer an anonymous, supervised paper trail.

Multiple checks on the trail with the ability to vote anonymously, in person prevents both fraudulent entries and coerced entries. Blockchain only covers 1 of the 2. And makes the coercion part more realizable.

Anonymous supervised paper trail is how many many states do elections. Scantron style ballots are simple and reliable.

I believe that the main objections to this approach are subconscious and unstated. Or perhaps people just feel silly saying "That isn't fancy enough!" and "But but but you have to use a computer to do everything!"
Some day someone will _surely_ find a real use for the block-chain.

Today is yet again not that day.

Prior to cryptocurrency there was no way to store your wealth online without relying on a centrallized third party that could theoretically confiscate it at any time. It's niche but it's not something to overlook.
Which cryptocurrency would you recommend as a reliable store of wealth?
I see what you did there! LOL
So you don't think this system provides any unique capabilities? I'm not aware of any other way to do some of the things proposed in the paper.
How does it improve on voting by making marks on paper?

as per https://news.ycombinator.com/item?id=18843548

Lots of ways. The votes are tabulated by a computer as they come in so results are instant. The totals cannot be changed or otherwise tampered with. There are no conflicts of interest in administering the election like what happen in Georgia with one candidate also being the secretary of state.
> results are instant

Irrelevant. The USA holds elections in November and seats the winners in January. What do you need instant results for? It's fine to take hours or even 48 hours after polls close in order to get it right.

The news media loves breathless real-time results because it allows them to lazily report on elections like they're sporting races, but I do not think that this is healthy for democracy. Reporting on poll numbers are in some places banned until the polls close, for good reason that it can influence the results.

> The totals cannot be changed or otherwise tampered with.

Wishful thinking. Fraud can still happen at the edges, when the data is entered. An electronic button press will always be less effort to fake in the regard that a mark on paper.

> There are no conflicts of interest in administering the election like what happen in Georgia

Non-sequitur. Elections should be independently administered regardless of medium, and person in this position would have opportunities to fix the results regardless of medium.

I hesitate to spend time refuting you because it's pretty clear you don't have a great grasp of the pros and construction of cryptography-systems, but here goes...

>Elections should be independently administered

The key word is "should." Your statement here shows that you either didn't read the paper at all or did and failed to grasp the most basic premise of the entire proposal; using the system elections are not administered by any party. Questions are posed and the system administers the vote collection and tabulating without human assistance or intervention.

>Wishful thinking

Not at all. The tamper resistance is mathematically proven. Unless an adversary controlled 51% of the network hashing power, it's impossible to change history. Of course the voter's client could be compromised but there are measures to get around such issues.

>Irrelevant.

I was responding to your original question of what does this do that paper can't. Now it seems you've moved the goal posts in order to argue with a strawman.

Look, I'm not a "crypto can solve everything" kinda guy, but it seems to me that if there's one area where blockchain tech shines it's in reliably collecting and retaining distributed information in an adversarial environment. That's exactly what voting is so I actually see this proposal as one of the few uses of crypto that's only argument for existence isnt "because distributed."

> the system elections are not administered by any party. > ... Unless an adversary controlled 51% of the network hashing power

Make up your mind.

> Not at all. The tamper resistance is mathematically proven.

Missing the point that this is irrelevant if the wrong data is entered.

> Now it seems you've moved the goal posts in order to argue with a strawman.

No, I'm arguing about what's best for elections. That IMHO is the most important point.