Some of it is counterintuitive. Most voting machines are, frankly, pretty crappy. We COULD try to make them better, but we'll never make them good enough to rely on. Fortunately, there are rigorous auditing methods that let us get reliable elections with UNRELIABLE machines.
So election security experts are focused less on improving the software and hardware (ultimately a fool's errand) and more on improving the election process (e.g., by incorporating audits as a routine part of the election certification process).
How to achieve this and whether it can be done depends on some specific properties of the voting system and the processes around it. (The important part is that it produce an archive of reliable paper ballots that reflect the voters' true choices).
Putting my (extremely well-worn) election security hat back on for a moment: it’s important to understand that the objections raised by 140-odd members of congress to various states’ election results were utterly and completely unsupported by serious or credible evidence. Period.
The violence at the Capitol yesterday will rightly be a source of outrage that we should remember. But we also must not forget the lies being told to cast doubt on the outcome by elected officials who should know better. They have forfeited our trust.
When I teach election integrity, I always take some time to discuss the fragility of the foundations of trust on which our democratic institutions are based. It’s always been hypothetical. Now it’s sadly far more concrete.
A somewhat under-appreciated and remarkable aspect of Elizabeth Friedman’s legacy is that she came to cryptology from a traditional literary background, just as it was shifting to being a chiefly mathematical discipline. Yet she continued to make central contributions.
She was also unafraid to call BS on her employer’s loony pet theory about hidden codes in Shakespeare’s writings.
The thing that gets most deeply stuck in my craw here is how the thousands of dedicated election workers who worked tirelessly to give us an amazingly smooth election at significant personal risk are not only being denied the thanks and recognition they deserve, but vilified.
To get a sense of what an accomplishment this election was, consider this short writeup from March on pandemic voting: mattblaze.org/papers/Emergen… . Almost all of the challenges came to pass (and more) and almost none of the suggested federal support materialized.
Essentially, most counties had to run TWO elections simulaneously - one in person, and one by mail, with no way to be sure in advance of how many voters would use which. And local election workers somehow did it. Really well. We own them a huge dept.
It's important not to allow the fact that high-level officials and media personalities have promoted these claims of election fraud to obscure just how utterly baseless (unsupported by serious evidence, easily refuted, or downright impossible) they have uniformly proven to be.
It's very easy to lazily conclude that because there are so many claims, and because prominent people have repeated them, that there must be SOMETHING to this. But there isn't anything to it. It's all just obviously fabricated nonsense.
A large volume of fabricated nonsense is still fabricated nonsense.
This would be a great move for Indian science. Though part of me understands scientific journal subscription fees the same way I understand paying ransomware.
For those not in science: authors and their institutions aren’t paid for their papers (sometimes it’s even the other way around). The journals don’t support the research in any way, but they insist on copyright transfer. Subscription fees are just pointless rent-seeking.