6 ms·

This has me wonder if highschools should start teaching the basic concepts of cryptography so that we eventually do end up with a common understanding of blockc

This has me wonder if highschools should start teaching the basic concepts of cryptography so that we eventually do end up with a common understanding of blockchains, password managers, passkeys, or any other technologies that we end up using in our day-to-day lives for crucial tasks.

runako 2y ago
> common understanding of blockchains This is funny because as a CS grad, I cringe about 75% of the time when blockchain enthusiasts make pitches that are oblivious to the workings of blockchains, the tech underneath, and their alternatives. If the blockchain community can't understand blockchain, it's going to be nigh impossible to convey comprehension to the general public. The general public generally just wants the authorities whose job it is to manage voting to do so in a competent manner. It's worth noting that there's really only been one candidate for national election in modern history who has called into question the fairness of our elections. (And then only when he lost.) Most of us understand that the folks who work for the Secretaries of State are generally doing the best they can with the resources we provide, and we don't want to provide more resources so they can do a "better" job.
echoangle 2y ago
It probably wouldn’t hurt, but you can’t really rely on Highschool education for election security. There are many people that don’t go to Highschool, and those that do probably forget half the stuff after a year. For an election, you want it to be basically obvious how it works in my opinion, anything that takes longer than 5 minutes to explain makes it easier to create doubt in the election.
tsimionescu 2y ago
That wouldn't help for this case. Even a PhD in cryptography and computer science doesn't help you in any way be convinced that a particular machine is securely counting your votes. If you want to be convinced of that, you have to audit the code and the hardware specs and the network code and everything in between to ensure that the system: (a) implements the claimed algorithms, (b) does so correctly and free of side-channeled attacks, (c) doesn't implement other things that can weaken the security after the fact, such as remote code download, and (d) has adequate physical protection to prevent hardware interference. And probably other things I'm not even thinking about. And all this work doesn't then help you ensure that another machine in a different jurisdiction, even one that is the same make and model, is also secure. Plus, every single person that cares about the vote has to put in this work for themselves: you can't "trust the experts" when the stakes are so high. I think this pretty clearly goes beyond what you could do teach a high-school setting.
- EGreg 2y ago
  This is wrong. You don’t need a Ph D or inspect code to know that your vote is included in a Merkle tree. And you can verify that the vote total matches what is in the Merkle tree for your district, and the national Merkle tree of districts. You can also verify that each voter was issued a unique token, which went through a mixer. About the only thing you can’t verify is that the agency giving out the token hasn’t been corrupted and gave a lot of voting tokens to fake people, or multiple voting tokens. That part (preventing sybil attacks) is why Voter ID laws exist throughout the world. But reducing the attack surface to widespread corruption issues involving voter registration, is much better than having those AND problets merely counting the ballots by hand, as when eg Al Gore lost to George W Bush in 2000. The other thing you can’t verify is that other people’s vote wasn’t tampered with — unless THEY report it. Which is why the voting system should require voters confirming votes from multiple devices that verify your cryptographically signed choices, eg vote on a laptop then scan QR code from that laptop with your phone and approve, just as you would with a web payment request in your bank app, crypto wallet or WhatsApp sign-in request. Because voting is not as valuable to people as securing their bank account, this requirement must be enforced on all voters. This way one company eg Google or Apple can’t spoof the interface.
  - tsimionescu 2y ago
    So you agree that you can't verify that the system has one and only vote tabulated for every person that actually voted, or that the vote they intended is the one that got counted. So, you agree that you can't trust the results of this election. Furthermore, if you check and find out that your own vote was incorrectly counted, you can't actually do anything about it, unless voter anonimity is not guaranteed: if you can't prove to an outside party what your real vote was, you can't pursue any legal action, you just know for yourself that the vote was rigged. And if you can prove to an outside party what you voted, that opens up a whole host of other attacks. So no, this is not even close to an acceptable solution. I'll also note that the Bush V Gore election issues were not caused by hand counting, but by machine counting as well. So, they should be taken as further proof that simple ballots and manual counts are the right way to conduct an election.

runako 2y ago

> common understanding of blockchains This is funny because as a CS grad, I cringe about 75% of the time when blockchain enthusiasts make pitches that are oblivious to the workings of blockchains, the tech underneath, and their alternatives. If the blockchain community can't understand blockchain, it's going to be nigh impossible to convey comprehension to the general public. The general public generally just wants the authorities whose job it is to manage voting to do so in a competent manner. It's worth noting that there's really only been one candidate for national election in modern history who has called into question the fairness of our elections. (And then only when he lost.) Most of us understand that the folks who work for the Secretaries of State are generally doing the best they can with the resources we provide, and we don't want to provide more resources so they can do a "better" job.

echoangle 2y ago

It probably wouldn’t hurt, but you can’t really rely on Highschool education for election security. There are many people that don’t go to Highschool, and those that do probably forget half the stuff after a year. For an election, you want it to be basically obvious how it works in my opinion, anything that takes longer than 5 minutes to explain makes it easier to create doubt in the election.

tsimionescu 2y ago

That wouldn't help for this case. Even a PhD in cryptography and computer science doesn't help you in any way be convinced that a particular machine is securely counting your votes. If you want to be convinced of that, you have to audit the code and the hardware specs and the network code and everything in between to ensure that the system: (a) implements the claimed algorithms, (b) does so correctly and free of side-channeled attacks, (c) doesn't implement other things that can weaken the security after the fact, such as remote code download, and (d) has adequate physical protection to prevent hardware interference. And probably other things I'm not even thinking about. And all this work doesn't then help you ensure that another machine in a different jurisdiction, even one that is the same make and model, is also secure. Plus, every single person that cares about the vote has to put in this work for themselves: you can't "trust the experts" when the stakes are so high. I think this pretty clearly goes beyond what you could do teach a high-school setting.

EGreg 2y ago
This is wrong. You don’t need a Ph D or inspect code to know that your vote is included in a Merkle tree. And you can verify that the vote total matches what is in the Merkle tree for your district, and the national Merkle tree of districts. You can also verify that each voter was issued a unique token, which went through a mixer. About the only thing you can’t verify is that the agency giving out the token hasn’t been corrupted and gave a lot of voting tokens to fake people, or multiple voting tokens. That part (preventing sybil attacks) is why Voter ID laws exist throughout the world. But reducing the attack surface to widespread corruption issues involving voter registration, is much better than having those AND problets merely counting the ballots by hand, as when eg Al Gore lost to George W Bush in 2000. The other thing you can’t verify is that other people’s vote wasn’t tampered with — unless THEY report it. Which is why the voting system should require voters confirming votes from multiple devices that verify your cryptographically signed choices, eg vote on a laptop then scan QR code from that laptop with your phone and approve, just as you would with a web payment request in your bank app, crypto wallet or WhatsApp sign-in request. Because voting is not as valuable to people as securing their bank account, this requirement must be enforced on all voters. This way one company eg Google or Apple can’t spoof the interface.
- tsimionescu 2y ago
  So you agree that you can't verify that the system has one and only vote tabulated for every person that actually voted, or that the vote they intended is the one that got counted. So, you agree that you can't trust the results of this election. Furthermore, if you check and find out that your own vote was incorrectly counted, you can't actually do anything about it, unless voter anonimity is not guaranteed: if you can't prove to an outside party what your real vote was, you can't pursue any legal action, you just know for yourself that the vote was rigged. And if you can prove to an outside party what you voted, that opens up a whole host of other attacks. So no, this is not even close to an acceptable solution. I'll also note that the Bush V Gore election issues were not caused by hand counting, but by machine counting as well. So, they should be taken as further proof that simple ballots and manual counts are the right way to conduct an election.