I would like to make a few additional points regarding the wires puzzle (a.k.a. the Chandrayaan Engine Room puzzle) in the column, and then switch over to some thoughts from an organizational perspective. The wires puzzle is an old one. There are versions of it all over the Internet. What is different here is the use of the Chandrayaan theme. Since this is happening aboard Chandrayaan, there could be relativistic effects that may need to be considered. The point of posting this puzzle using the Chandrayaan theme is to say that, sometimes, old problems can be looked at in new ways.
That is, one interesting way to think about the problem is to ask if the solution to the problem changes if we have relativistic effects. If the rate at which the wires burn changes over time, then you suddenly have a more interesting problem. The point is that if we shift the 'frame' of the problem, then sometimes, the problem itself changes. (See Steven Pinker's discussion of some of these 'framing' effects in Chapter 5 of his book "How the Mind Works".)
We are planning on a puzzle in Finance in a future column. With Finance problems, there may be other assumptions such as time preference of money. Milton Friedman talked about problems and assumptions in his famous paper "The Methodology of Positive Economics". As Friedman put it, theories almost always make assumptions, and so the solving of a puzzle is a Socratic process in that the framer of the problem and the anwerer must agree on a certain set of assumptions. But beyond agreeing on the set of assumptions, nothing more is required.
Now, puzzles are often used in interviews, and one person even termed his interviewing process Socratic. I do not think, however, that a Socratic approach to puzzles is a good one. It seems to be a rather mistaken approach. From what I have seen, the dialogue that ensues after a puzzle is proposed is almost never about clarifying assumptions behind the problem. The dialogue is about the interviewer providing the interviewee subtle clues to solving it. How many clues the interviewer provides the interviewee depends on how much the interviewer likes the interviewee at first blush. And that's not ideal if we want the organization to pick the best people since first impressions can often be mistaken. There is a problem here that I am calling your attention to and it is an organizational one. Most interview puzzles are quite sufficiently specified, and so the requirement these days that interviewers engage in a dialogue with the interviewee seems rather wrong headed. It ought to be sufficient for the interviewer to clarify assumptions underlying puzzles, and then let the interviewee figure out the rest on his or her own.
The whole point of puzzles is to have an objective means of analyzing candidates. There is much subjectivity in almost every other part of the process. Companies really ought to change the way they interview candidates. Why companies continue to do interviews they way they do is, of course, a whole different ball of wax.
Update (June 18): This is a very technical post. You can safely skip this post and still enjoy the puzzles in this column. You can also assume that there are no relativity effects in coming up with a solution.