Iceberg problems
other posts

Tiny little problems that you can make progress on without them getting smaller.

Today I had one of those experiences. I wanted to tell someone, “‍you know the ones, where you notice this tiny little problem and you know how to fix it and eight hours and dozens of fixes later you realize you’ve fixed all bu a tiny little part of what was actually a really complicated problem.‍”

And then I realized that of all of the many aspects of my life, I can only ever recall having these experiences when engaged in two tasks: programming an algorithm of my own design (and thus finishing the design as I go) and writing a proof for a theorem of my own devising (and thus refining the theorem as I go). I can’t recall having it when writing papers, handling logistics, preparing lessons, designing RPG modules, illustrating, cleaning, fixing physical objects, working on family history, working on standards design, or any of the many other things I do. And most people never prove their own theorems or implement their own algorithms. Meaning this so-common-it-seems-it-ought-to-be-a-unversally-experienced-archtypical experience for me is likely one that most of my friends, family, and readers have never experienced.

The most similar experience I have is with office politics. I enjoy the process of understanding and attempting to influence a society, but I find that what seems like a simple change often takes years to implement. You start by finding out who has the authority to change it, and then you talk to them and find out there is another player you need to consult, and that player mentions a third, and then someone suggests a vastly different perspective that also needs to be considered and as you ask about that you learn of something else and soon years have passed as you have explored this vast web of social interactions. There are brief periods when you think “‍wow, this is hard‍” but there is almost always one or two simple next steps, which once taken bring up one or two more, and so on. The main difference between this and the other two situations is the scale of the individual steps. Politicking spans days at a minimum, meaning it is always interleved with other activities. Coding and proving spans minutes, and one bundle of minutes can run into another with no break, leading to hours flying by uninterrupted and unnoticed.

There are some more common experiences that are a little like what I mean. Explaining a job you know well to a new hire can be like this: you list the three things you do, and then you realize there are four more you forgot to mention, and then a few more, and so on for hours or days. It’s also a bit like having a passenger in your car give you verbal directions into a complicated maze of cul-de-sacs: you always know the next step and assume it will be the last one, but then there’s another, and another…. But these are getting farther from the experience still.

Trying to metacognate, I think the key aspect of these activities is that they involve deeply nested but shallow concepts. The next step really is a small next step, easy to grasp and identify as next and not that hard to implement, but once you take that step you are now a bit more deeply in and have the perspective to see there is another individually small piece to tweak. They have the unusual characteristic that you can’t see the journey, only the next step or two. You might know, from experience, that “‍things like this are often more complicated than they seem‍” but it’s just abstract, and sometimes (even often) they aren’t: you do the small step you see and it works and you move on.

Because I find myself wanting to discuss this common experience of my life, and because I don’t seem to have a good word for them already at hand, I propose we call them “‍Iceberg Problems‍” and working on them “‍Iceberg Mining‍” from the oft-analogized aspect of icebergs being mostly hidden beneath the sea. You see a bit of ice floating there, you remove the part you can see, and up pops the next bit.

I spent most of today iceberg mining. So much so that the little task I thought I’d just hack out before breakfast ended up taking me until dinner time to finish.

One more point about iceberg mining that those that have not done it might not realize: it is fun. Almost intoxicatingly fun, sometimes, leading to what I once heard If memory serves, I heard Chuck Knutson attribute this to Steve Balmer, though I have no documented source of either Knutson nor Balmer saying this. described as “‍the happy coder bathroom dance‍” where iceberg miners would rather continue mining than attend to their body’s basic needs, squirming in their seats in an effort to make the urge to urinate dissipate.