7 ms·

> And so bending over backwards to absolutely ensure there's no reasonable doubt whatsoever There is no reasonable doubt. To the extent that any doubt remains,

> And so bending over backwards to absolutely ensure there's no reasonable doubt whatsoever There is no reasonable doubt. To the extent that any doubt remains, it is only in the mind of people who wouldn’t be swayed by facts. As someone who have spent the Trump years learning about epistemology (how we come to believe the things we do), I have realized that motivated reasoning will always blind humans to the truth. The truth is that elections require some amount of trust. It’s not possible to allow every voter the ability to see and verify each of the 162million ballots and no election would work if that was a requirement.

oleril 3y ago
Bitcoin the most secure public ledger in the world proves that every participant in the blockchain can verify and see billion+ valid transactions. Trust in fully decentralized and transparent elections is not a technical issue!
- thephyber 3y ago
  Lol, you again. The BTC ledger is pretty good, but it is fallible —- mostly because a human has to control the keys and the UI apps that make using the wallet for transactions are far less secure than the core BTC protocol. Good luck teaching a 90 year old person who hates computers how to vote using their BTC wallet. Also, the security of BTC isn’t infallible. Once someone finds a weakness in the system, they will quietly abuse it as a 0-day for a while before anyone notices it. Elections should be robust against any single point of failure, and they are. The BTC wallet is until it isn’t.
- bcrl 3y ago
  Bitcoin sustains about 7 transactions per second. In 2022 there were roughly 168 million registered voters in the United States. It would take 24 million seconds to publish 1 transaction (vote) per voter using a bitcoin style blockchain. The math doesn't work. Bitcoin is not a solution to this problem.
  - thephyber 3y ago
    Throughput isn’t fixed across all blockchain implementations. It’s possible to fork/side chain BTC to accommodate far more transactions per block, thus drastically increasing throughput. I still think _ergonomics_ of making sure every eligible voter can use the system is far more important that making it {immutable, distributed, transparent}. Also, the immutability property is a double edged sword as someone losing control of their wallet secret key would lead to a leak of everyone they voted for. Or worse yet, if voter ID is tied to wallet ID instead of government-issued ID, one organization could go around stealing private keys in order to steal those ballets.
somenameforme 3y ago
Ballot verifiability is not impossible by any stretch of the imagination. A person fills out a ballot, that ballot is [optically] scanned and then automatically parsed a la scantron. Both the ballot and the resultant scan are published side by side. Optionally, add an ID on each ballot and people can not only look at all 162 million ballots, but even verify that their ballot properly scanned, published, and counted. The only argument against this is that ballot verifiability is a bad thing, because it might enable e.g. vote buying or whatever. Yet somehow people were perfectly fine with there being more than 26,000,000 votes made by mail in 2020 in which such threats are dramatically greater. The nice thing about this solution is also that there's no 'classified' data. The complete transparency of it all would make any sort of stuffing or vote manipulation effectively impossible, which means there's no need for ultra-sophisticated systems. Plain old 1990s tech of digital scanners and scantron would work just fine. Ballot observers would need to do little more than count the number of people coming into a station and ensure that matches the total number of ballots eventually published by that station. What's not to prefer about this relative to the status quo?
- thephyber 3y ago
  Spoken like someone who has never actually looked at how an election works in a county in the USA, doesn’t care about the actual legal requirements, seems to ignore all concerns for accessibility (eg. Scantrons are terrible for old people with bad/no eyesight, have motor control issues with hands, etc), doesn’t know what scantrons actually did with the scan results, hasn’t thought through the complexity of ID in the USA (we don’t have a single ID number and the alternative is to store photo ID cards for everyone to see and collect) and doesn’t have a plan for absentee voters or military deployed voters. And that doesn’t even address any of the human factors (eg. people lose their ID, they lose their ballot, ballots get crumpled by voters and stuck in the scanners, electricity isn’t up 99.999% of the time, etc). Elections are hard for a reason — the requirements are complex. Everyone wants 160million+ votes to be collected, counted, and results published by 9p on election day. Everyone wants the registry to work perfectly and wants the convenience of not having to travel far to a polling place. In practice, these create a huge number of hurdles that election officials must navigate, plan for contingencies (eg. the hurricane immediately before the election in Florida 2022). Maybe my best advice to you is the snide “don’t reinvent the wheel in Hacker News comments, but rather get involved as a polling place volunteer next year, then see all of the edge cases in action”. Then don’t propose your MVP v2 to me, propose it to your county election board. Only you will be ignored because elections in the US have 250 years of learning from the previous errors, but you seem quite content to ignore all of those hard-earned lessons.

oleril 3y ago

Bitcoin the most secure public ledger in the world proves that every participant in the blockchain can verify and see billion+ valid transactions. Trust in fully decentralized and transparent elections is not a technical issue!

thephyber 3y ago
Lol, you again. The BTC ledger is pretty good, but it is fallible —- mostly because a human has to control the keys and the UI apps that make using the wallet for transactions are far less secure than the core BTC protocol. Good luck teaching a 90 year old person who hates computers how to vote using their BTC wallet. Also, the security of BTC isn’t infallible. Once someone finds a weakness in the system, they will quietly abuse it as a 0-day for a while before anyone notices it. Elections should be robust against any single point of failure, and they are. The BTC wallet is until it isn’t.
bcrl 3y ago
Bitcoin sustains about 7 transactions per second. In 2022 there were roughly 168 million registered voters in the United States. It would take 24 million seconds to publish 1 transaction (vote) per voter using a bitcoin style blockchain. The math doesn't work. Bitcoin is not a solution to this problem.
- thephyber 3y ago
  Throughput isn’t fixed across all blockchain implementations. It’s possible to fork/side chain BTC to accommodate far more transactions per block, thus drastically increasing throughput. I still think _ergonomics_ of making sure every eligible voter can use the system is far more important that making it {immutable, distributed, transparent}. Also, the immutability property is a double edged sword as someone losing control of their wallet secret key would lead to a leak of everyone they voted for. Or worse yet, if voter ID is tied to wallet ID instead of government-issued ID, one organization could go around stealing private keys in order to steal those ballets.

somenameforme 3y ago

Ballot verifiability is not impossible by any stretch of the imagination. A person fills out a ballot, that ballot is [optically] scanned and then automatically parsed a la scantron. Both the ballot and the resultant scan are published side by side. Optionally, add an ID on each ballot and people can not only look at all 162 million ballots, but even verify that their ballot properly scanned, published, and counted. The only argument against this is that ballot verifiability is a bad thing, because it might enable e.g. vote buying or whatever. Yet somehow people were perfectly fine with there being more than 26,000,000 votes made by mail in 2020 in which such threats are dramatically greater. The nice thing about this solution is also that there's no 'classified' data. The complete transparency of it all would make any sort of stuffing or vote manipulation effectively impossible, which means there's no need for ultra-sophisticated systems. Plain old 1990s tech of digital scanners and scantron would work just fine. Ballot observers would need to do little more than count the number of people coming into a station and ensure that matches the total number of ballots eventually published by that station. What's not to prefer about this relative to the status quo?

thephyber 3y ago
Spoken like someone who has never actually looked at how an election works in a county in the USA, doesn’t care about the actual legal requirements, seems to ignore all concerns for accessibility (eg. Scantrons are terrible for old people with bad/no eyesight, have motor control issues with hands, etc), doesn’t know what scantrons actually did with the scan results, hasn’t thought through the complexity of ID in the USA (we don’t have a single ID number and the alternative is to store photo ID cards for everyone to see and collect) and doesn’t have a plan for absentee voters or military deployed voters. And that doesn’t even address any of the human factors (eg. people lose their ID, they lose their ballot, ballots get crumpled by voters and stuck in the scanners, electricity isn’t up 99.999% of the time, etc). Elections are hard for a reason — the requirements are complex. Everyone wants 160million+ votes to be collected, counted, and results published by 9p on election day. Everyone wants the registry to work perfectly and wants the convenience of not having to travel far to a polling place. In practice, these create a huge number of hurdles that election officials must navigate, plan for contingencies (eg. the hurricane immediately before the election in Florida 2022). Maybe my best advice to you is the snide “don’t reinvent the wheel in Hacker News comments, but rather get involved as a polling place volunteer next year, then see all of the edge cases in action”. Then don’t propose your MVP v2 to me, propose it to your county election board. Only you will be ignored because elections in the US have 250 years of learning from the previous errors, but you seem quite content to ignore all of those hard-earned lessons.