It doesn’t solve it, it just moderates it. Thanks for keeping me honest here. I normally would say that, but I was a little...One way that it talks about is that you have to, when issues are contingent on each other, you have to serialize the decisions on them.

At the end of each chapter, there’s typically a few paragraphs about what you could do if you wanted to get rid of that paradox. It always comes down to, "Well, if there were only binary contests." It’s the same thing.

When you read that it’s fascinating. It was like a total validation of RSV to me, because I only read it afterwards. Every chapter talks about a different electoral paradox like Arrow’s theorem. There’s all these different things. More generally they’re called "aggregation paradoxes."

He backs random-sample voting, by the way. He wrote this beautiful monograph called "Electoral Paradoxes." You’ll see him on our RSV website.

For a whole week, we had stuff in churches and we stayed at the monastery. Anyways, I can tell you that the electoral rule expert or whatever, expert in electoral rules is Hannu Nurmi. Everyone there deferred to him. He was the guy. He had a bibliography of several thousand ...

He’s a Finnish, emeritus professor of electoral systems. I was once in a conference of about 25 people from around the world that were the top people in the world of electoral systems design and it was in Erchie Italy, in Sicily, in this little stone village.

They did it by bringing every question down to a binary vote. Do you know this guy, Hannu Nurmi?

It was a little bit, would you call it after-the-fact kind of control? They let legislators do stuff, but in case it violated the constitution, there was this after-the-fact way to prosecute them or probably invalidate the laws.

They add the vowels. That’s the Internet of those days. They basically empowered citizens to be able to read and write and then they control the whole thing, in effect through RSV.

It’s actually the only way so that in 140 years, they created all the elements of Western civilization really, from the musical scale to philosophy and medicine and particle physics and the rest of it. They had the first phonetic alphabet. That’s my contention anyway.

You could be put to death as a volunteer legislator, for proposing a law that was unconstitutional. That shows, to me, and you can argue, that they really did run their whole democracy that way. That’s why I call random-sample voting a proven way to run a group of citizens, ...

Then the two would argue. If the jury found that it was, in fact, unconstitutional, then the two would propose a punishment.

It’s kind of pretty close to the Greek jury system except they didn’t really use it to create legislation. They used it to, in effect, approve legislation. Anyone who proposed legislation that someone else thought...Sue them. Any citizen could sue a legislator -- sorry, I have to say this -- ...

You can use the tournament-style thing to make sure that the stuff that gets put to a real vote -- and people have to pay the most attention to -- is the high priority and reflects all the best input and then you decide whether you agree with it or ...

The tournament design people in sports, they really worry that there would be an error and the absolute best team wouldn’t win because they got eliminated. Second best wouldn’t win, that’s the thing, and that kind of thing, wouldn’t get the second best title.

If you look at tournament design, it’s a thing to make one that’ll give you top 10 for sure, but it’s pretty easy to give you the top 10 probably and then you don’t really have to pay too much attention.

The tournaments don’t have to use the full-out sample size in the lower rungs because they could be a slightly larger air probability there, but when you get to the top, it’s a way to really get high probability, the best few...

It’d be way too expensive. It just wouldn’t work. That’s like this. You want to allow as much and as flexible and really neat brainstorming and Post-Its and everything. If some other guys have a better idea and maybe they are proactive or not, who cares? It’s all good.

That final stamp of approval. Really, no one cares that much about all the rest of the paper in that plant. If they didn’t do it that way, it wouldn’t work. It wouldn’t be anywhere near as secure as it is.

Then they get sliced up and bundled and they mail them -- believe it, it’s the weirdest thing -- they mail them by registered mail to the banks.

There’s basically one machine in there that puts the seal on the bank notes. Every single bank note gets that little seal with the serial number and there’s just this one machine that does it. It’s not even sheet by sheet. I think it’s part sheet by part sheet.

They’ve got all these machines and everything, but then there’s this one room, it’s not really much bigger than this room. It has a taller ceiling because there’s a catwalk around where you can watch it.

They took a photo of me leaning on a palette full over $100 bills. It was pretty cool and everything. The interesting thing about that place is it’s a huge facility. It’s a big industrial place, they’re printing bank notes.

The only good way to do it that I know of and works really neat because it’s super definitive. When I was in electronic money, for a photoshoot, they took me to the US Mint in DC, I think. The Bureau of Engraving and Printing, where they print the bank ...

Letting a thousand flowers bloom, like liquid democracy, all these deliberation forums, expert lobbying, all this junk as input but then using the RSV as the final imprimatur that this is violated by the general public in view of all this analysis and argument that’s been placed.

Also, the jury sizes are pretty large compared to...The minimum was I think 300. It’s very much like RSV. Then the interesting thing...

I’m glad you know about this. They’d go out with those disks. They weren’t allowed to talk to each other. Some of them had holes, some not. They put their their arms in, and they released them. It’s almost as secure as pope elections for the last 600 years or ...

The rows, the IV parts, and the tribes, the districts that were stripes, the whole thing. Then the jurors would go in. They weren’t allowed to talk to each other. The two contestants would argue it out a little bit. Then the jury would leave. They’d have the disk.

Oh. They would show up at the jury place. The archon or however you pronounce that, the armed guy there running the thing, would pour a jar of or urn of marbles. They were mixed alternatively black and white. They would go down this slot. Do you know about that?

You know, it’s in the article, the way the ancient Greek juries worked. I didn’t realize it when I created this, but it turned out I basically recreated the exact thing they had. Did you know that when the...Jurors during the Classical Era had to be of a certain minimum ...

The more you have...That’s the beautiful thing. It’s not like election server that has the secrets. It could be penetrated. It’s online. No, there’s just a bunch of these things. They collect it. It gets posted. It’s pretty complete and pretty tight. It could be used for a lot of ...

Yeah, maybe other people would do it just to prevent DDOS because that’s of the unsolvable problems according to the anti-online voting people. "Oh, well, wouldn’t want DDOS. What if the election were delayed? What if the whole rule were delayed?"

There is no computer that always has secrets in it. It’s only reconstructed during those meetings. Then the bulletin boards don’t know enough to do any harm. They can in fact be run by different interested parties. It’s pretty robust.

Next time they come together, we call those "meetings," they put in the thumb drive and the passcodes. They’re off the races again to do the next step, which is generate the PDF ballots. Then they sit there while they print those out. It’s neat.

That’s why it’s all divided into those steps. It recreates the secret state from that. It gets the randomness from those keys. Then it does the one state. Then it posts that step nine. Then they witness its complete erasure. They go home happily, but they save this enscripted state ...

The election officials put those in, and they boot up this virgin desktop machine. Then they each type in their passphrase, which is a cryptographic key-length string, and then it starts up and it runs step one.

What we did to make it practical -- the election officials were happy with it and support what we need. We have a computer that we boot up from a thumb drive that has the bootable image on it, and it boots up.

But no election official’s today [laughs] going to run their own secure computer, and if they did they might not do a good job. You can’t use all that fancy public key stuff, because the public’s not going to understand it.

The way we did it in Takoma Park, Maryland...A cryptographer would just approach this typically and say, "Well, we’ll use a lot of public key signatures and multiparty computation," actually something I helped invent. "There’ll be boatloads of computing. Then, every election official will have their own secure computer and ...

It’s just a great thing that you can make something that’s understandable that has that level of security. I want to just mention one other security trick we do. You might say, "Well, that’s great, but the election officials have all these secrets that they use to generate this, and ...

Thank you for for taking the time to understand.

It gives no clue about the others.

Actually, there’s 10,000 codes that would yield that same hash that they’ve stored. The chance that they could guess what the valid vote code is insignificant. They could check, and then, they’d give you the standard user feedback, but they can’t falsify them. Of course, they can’t swap them knowing ...

In fact the vote codes are actually much longer numbers, but the voter bulletin boards, they only have some digits of the hash of the vote codes so that when you type it in, they could recognize that it’s valid.

Thank you. I appreciate that enormously. I’m going to do that.

Yeah, and the fact that these particular ones are all miraculously in order is for convenience in illustration, as are these repeated digits [laughs] and all these other tricks to make it a little easier to understand.

There are two more. Now of course, they wouldn’t all be in this canonical order. I show that as an example, you can do that.

It’s just that I didn’t...I think in this example, I showed all four rows for this example, double-ballot 100, is that correct? I showed all the cases that, but for the other one, I didn’t want to make it the big illustration...

It’s always kind of...

With the tables that are open where you can see the votes, those are different tables, we never see what the summands are in those open tables. Do you see what I’m saying?