How It Works
This puzzle was based on a toy Andy Rich had as a kid: a grid of twelve cubes that had six unique jigsaw puzzles on them, one per face. That idea was then expanded upon. There were two twists here: one, to discover that, aside from one face, none of the other pictures were the expected 4x3 shape (you had to "think outside the box!"). Two, to figure out how to extract information.
The six pictures to assemble were: a story, a gun, a bottle of scotch, a noose, an ashtray with cigarettes, and a picture of a city.
Being text, it was expected that the story would be assembled first. (This also happened to be the only puzzle that assembled into a 3x4 grid.) When assembled, the message said: "Typical day: fire a gun, drink some scotch, tie a noose, smoke some cigarettes, leave the city, and watch the stars."
This step provided a few vital pieces of information: a hint to the images on the other five sides (to aid in assembling them), a hint about "watching stars", and an ordering. The other pictures, when assembled, looked like this:
Extracting the Message
The hint about "watching the stars" comes in quite handy when extracting information. Looking at the bottom of the box that the cubes came in, it's clear that you're expected to extract information:
The trick, in this case, is to recognize that the cube with the star on it always goes on the square in the box with the star. When the picture is placed in the box with the star cube covering that position in the box (with both the picture and box at obvious orientations), you can extract the message pieces. For example, in the picture below, the scotch bottle picture gives you "CU":
The ordering for the message pieces is provided by the story in the beginning: gun, scotch, noose, cigarettes, city. When you extract the full message, it reads: LOCUTUSOFWHO; the answer to which is BORG.
The images were all produced by Andy Rich, who has virtually no artistic talent. However, he is able to trace, and so these images were traced in Microsoft Expression. Some source images:
(If you're curious, that cityscape is Chicago.)
A homemade C# program was used to chop the full images of the puzzles into their small pieces, and these were assembled into the documents for the cubes themselves in word - one picture per cube. The paper that the cubes were printed on was purchased from SlidingCards.com and printed at Kinko's. (In general, the printers at Microsoft were not precise enough to do a good job printing.)
This puzzle was fabricated and ready a few weeks prior to the actual event, and sat in storage. However, two days before the event it was suddenly remembered that no one had put the "answer" paper in the boxes that contained the cubes, so the puzzle would have contained cubes, but with absolutely no way to extract the answer! (In other words, this puzzle was nearly a disaster!)
The only real difficulty with this puzzle was that it required a little odd knowledge: knowing who Locutus was. In retrospect, this was a little unfair to some teams who didn't have that knowledge - we tried very hard in the pregame to make sure you knew all the colloquial knowledge you needed, but this was overlooked. Sorry, guys!