With over six million views in six days, OK Go's video for "This Too Shall Pass" is the latest in their unprecedented string of runaway YouTube hits. Lead singer Damian Kulash shared OK Go's video philosophy—and history—with Gizmodo.

Q: At this point, OK Go may have the best track record of anyone at creating these incredible viral music videos. Why are yours able to stand out?

A: I think it has become increasingly clear to us, as we have chased our most exciting ideas, that there's been a dissolution of the distribution system for music. That seems really depressing when you see that records aren't selling, there's no way to make any money, the system's falling apart. But if the system itself is falling apart, then so are the rules wrought by it. Videos evolved in this very restrictive environment of MTV. There were only maybe 100 that would play at any time, and labels weren't willing to invest in them. So now that the system is falling apart, there's also no reason to stick within the confines of the definitions that were built up during that system. This sounds heady and pretentious, but it means for us the ability to chase our most compelling ideas. We don't have to think so much into the box of "Will this song work on this radio format?" There's an infinite world of possible audiences out there for whatever you're making now.

It's not like we sat down one day and the brain trust came up with idea for "This Too Shall Pass." Tim [Nordwind, the bass player] and I have known each other since we were 12, and it's always been the animating passion of our lives to make fun projects together. Everything from making home videos to recording songs. So the fact that the band got signed and gets to make records is all well and good, but that's all just a part of our creative relationship. Now that we have an outlet for these other things, all the better. The video for "A Million Ways" [below] was originally just a practice run for a live show. When that catches fire… We're now in a position where we don't see restrictions on what we can do at all.

Q: So "A Million Ways" wasn't even supposed to be a video at first? How'd you stumble onto that dance?

A: Before we were even signed, we were all living in Chicago and there was this incredible public access show called Chic-a-GoGo. It's like a lo-fi Soul Train. You bring a five year old in and an art student with a gorilla neck and everybody has a dance party. We only had one song at that point. We got a chance to perform there, which was great, but it was so low budget that they couldn't record our audio. We decided if we're going to lip sync let's swing for the fences, and we came up with this totally ludicrous dance routine to the only song we'd at that point recorded ["C-C-C-Cinnamon Lips"].

Tim worked at NPR at the time, and Ira Glass was a fan. He took us on tour as his opening act for "This American Life," and we kept the dance routine.

Rock shows are such a known quantity. The band does this, the audience does that, and there's a particular range of emotions people go through. But when you bring something people don't expect, it really shakes it up and is very different and weird and fun.

As for the dance for "A Million Ways," we'd come out with our second record and we didn't want to do the same dance that we'd done for our first. My sister choreographed a new one for us, and we worked on it in our backyard. The video was a practice tape, but there was something so funny and awkward and weird about it that we just sent it around to friends. Then it suddenly had 500,000 hits, which was more records than we'd ever sold.

I truly and honestly did not believe that numbers close to that video's were achievable again. A lot of it was dependent on YouTube being brand new at the time, and people discovering the service when the video came out.

Q: Do you feel pressure now for every video to go viral? Especially one that took as much time and effort as "This Too Shall Pass"?

If "This Too Shall Pass" could have a broader footprint than "A Million Ways" or "Here It Goes Again" did, that's great. But that's definitely not our intention. From our perspective, the upshot of these things being successful is the ability to do a lot more of them in the future. We've done a lot of videos in the last few years. I'm definitely happy with the video of "WTF?" and this latest one, but when these videos do well it makes it so much easier to get the other ideas we've conceived done. Saying "I'd love to do this thing [in a video] with six cars" is tough, but now it's more likely that someone will actually give us six cars. It's less that they're designed for viralness and more that the operating principle of our creative life is chasing down those ideas.

Q: Where did "This Too Shall Pass" come from? Do you consider it a continuation of your previous efforts or a jump forward?

A: "This Too Shall Pass" is a combination of a bunch of things. Making the treadmill video ["Here It Goes Again"] and the wallpaper one ["Do What You Want"] after that, I just got really obsessed with these contingent systems. Looking at choreography not as dance or movement but as a performance or a system that requires lots of disparate elements to work in perfect synchrony, or sometimes imperfect semi-synchrony. I was thinking a lot about loosely choreographed pieces. What sort of systems can you do that aren't specifically dance, but you get the effect that the whole is greater than the sum of its parts, because everything works just where it should. Rube Goldberg machines are also, I think, universally magical.

Our label, bless their moronic hearts, was given our record nine months ago. It kept getting pushed back. We basically wound up with several months of our lives to just get in trouble. If we'd had to go into promo land and get on tour we wouldn't have time to do this kind of stuff. Basically I got home when the record was done and wrote down my dream list of videos. This whole project started with a two-paragraph description that I put down online as a job post, essentially. I asked for two creative engineers, because I figured that's about what it would take. Two engineers, and a couple of months. It ended up being more like 60 engineers, and five months of work.

Q: The set looks like a walking death trap. Did anyone get hurt, or were there any close calls?

The camera man was actually hit by the giant blue barrel that falls from the ceiling. You see the camera takes jolt at the end, right around the time the airplanes take off. That's the big blue barrel running into him. Otherwise, there were a few bumps and scrapes and bruises. Brett got hit by the bowling ball when it didn't stay on the ramp once. But none of the super dangerous things every hurt anyone, that I know of.

In terms of putting ourselves in harm's way, what makes these things exciting is the experience quality of it. The essential element of this would be lost if we made a film that depicted all these components but didn't actually have them. I can think of other music videos that show Rube Goldberg machines, but they're all carefully edited things. It loses the idea of being there for the people doing it or the people watching it. I pushed the design team to make the ways the machine treats us stranger, rougher. I was hoping the part where I get rocketed across the room would be a catapult. The professional circus riggers who set that up said we couldn't do that. I was pretty insistent, but they were very clear that no, making you airborne is going to hurt you. And I was like, don't people do this all the time? And they said sure, there are stunt men who train for years and/or do this with a lot of CG. I wanted to do it, but apparently this is as dangerous as it gets.

Q: Wait… you had circus riggers on-set? What other professionals came together to help build this thing?

A: It was such an incredible group of people. The doors that fall at the end were designed by a rigger/builder guy who everyone called "The Pirate." His mains source of income is working on longships, so he's actually literally a seaman. The person who painted them is the guy who designed the most recent Coke bottle. It was a crazy group of people. The reason we got that spread is we didn't walk into USC and ask for their brightest engineers. We posted this stuff on the MindShare Labs list. I think they're called Syyn Labs now. They're basically a community of nerdy, creative folks in Los Angeles. Anybody who was wont to go to a lecture series as a drinking venue had access to this. Basically anybody who sees the smart/fun/creative side of engineering.

Q: Why such emphasis on "old-school physics" and practical effects instead of CG?

A: On the basic level, this whole project is only exciting because it is real. It's not a labor of love for anyone to go make a commercial. This is an art project for all these people. If it ain't the real thing, it's not worth it. They're not there to make a video that promotes a band. They're there to make this awesome project. Any time someone suggested a way to do something easier, I gently pushed them away from it. What makes Rube Goldberg machines so universal is very hard to describe and very easy to lose. If you make it failproof, the thing completely loses its magic.

Q: Would you say that's how you've historically approached your videos?

A: Across our videos in general, it's not really a requirement but it's something that attracts me. I once wrote out a list of 20 things that make a good video. One of them is that it's something that anyone watching could, with enough time, have done themselves. Treadmills and choreography and all the things in "This Too Shall Pass," none of those are specialized access. A broken piano costs like 70 bucks. It's not like you have to be an engineer to get that.

Look, we were working with engineers from NASA. Three people who worked on the Mars Rover worked on this machine. And it was wonderful getting people to stop using the specialized part of their skill and get them to use the inspired part. A lot of times I had to explain what "magic" was and what they weren't allowed to do. You want to use optical gates? Okay, but it has to be followable for the audience. What about lasers? You can't use something from your lab you worked in, but you can use a laser pointer from a gas station. What if you dissected a Blu-ray player? Fine, but only as long as people can tell it's from a Blu-ray player. You'd be surprised how much communication it takes.

Q: Any parts in particular stand out where you could've been spared a lot of trouble given a CG or manual assist?

A: Almost every point in there could have been cheated. There's no way to cheat the table I'm sitting on in the beginning. I suppose you could maybe put together that machine and then animate the balls but that would've been incredibly difficult. Almost everything else would have been a lot easier with a manual cheat or CG cheat. The timing on everything was critical.

Take for instance the sunrise contraption, the umbrella that comes up as the sun. The timing delay between the sun coming, the flowers coming up, and the birds coming down—we could have just triggered all that stuff electronically or manually. The way it was actually done is changing the fulcrum of the 2x4s that those things were spinning on, so the weights on the end would spin around more slowly. A hammer gets hit on the fulcrum on the back, and by changing where that hammer was, you change the delay until the release of the flowers. That kind of stuff, there's no reason we couldn't have cheated all this, but the 60 people who built this thing wouldn't have had the challenge and the satisfaction of the finished product.

Q: So what's next? Do you feel pressure to keep topping yourselves?

Mostly I'm just excited because I think this makes it more likely that we'll be able to do more in the future. Finding people who will help us pay for some of these things should be a lot easier right now. And finding collaborators. As wonderful as the team was, there's no way that those people—no matter how compelling an internet posting I'd put up—there's no way they would've signed up to do this if we hadn't already done the treadmills. The success of any particular project is that rather than lifting the bar and creating pressure to come up with new ideas, it opens you up to more and more of them. Now it's more likely that when we call to find an anti-gravity chamber in NASA, it'll happen.

Q: Ha. Is that something we can expect to see at some point?

A: Oh, man. Weightlessness would be the final frontier, I think.