974: The General Problem
This comic features Cueball sitting down to a meal and requesting that an offscreen person pass him the salt. The offscreen person (OSP) then proceeds to solve the problem...generally. Cueball's implied/specific request was "Can and will you pass me the salt immediately?" However the OSP begins to solve the salt problem generally, without regard for time, context, the specific heat of the meal, or what some would call common sense. For the next 20 minutes, while Cueball nibbles at his bland, cooling meal, the OSP works on a system that will pass condiments more quickly than possible by the OSP. The problem, obviously, is that developing and then using the system is much slower than just manually passing the salt.
In the title text, Randall states that when someone uses a lot of time to do an easy task right in the present he considers them a perfectionist with no ability to prioritize (since they are unable to do something inefficiently even if it satisfies the task at hand much quicker). Yet if a perfectionist had done something right in the past he would be impressed and consider them a master artisan of great foresight.
The OSP's argument for taking so much time is that in the end if he were to add up all the time he saved by no longer needing to pass condiments it will equal more time than it took to build the system, and thus in the long term he will have saved time and solved the general problem of passing condiments. This could also be spread out to many people all over the world, and thus save a billion times as much time, if a billion people in the world have both condiments and an adequate food supply to use them on...
This situation would be akin to a major website performing maintenance during peak hours instead of waiting till traffic was lower. From the perspective of the person trying to check their email, the upgrade would seem ill-timed and unnecessary. However, if the person were to instead log in not till the following day, he would think the upgrade was masterful and full of great insight.
Put simply, the comic draws attention to the fact that humans are very adept at things that can be difficult to teach machines. We learn as children how to pass arbitrary things to someone who asks. However, to design a 'system' to do this, can be quite difficult - dealing with recognition, shape, size, temperature, etc. There is an engineering joke here that novices often attempt to solve general systems problems that seem trivial at first sight, but can end up taking years to solve. The superficial joke is that a human would have to think at all to do this, let alone 20 minutes.
The title text carries perhaps the more important point, for which of course this case is a trivial and hence humorous example: most pure and some applied research does not look like it is a reasonable use of one's time. Therefore, only in retrospect, when something has in fact resulted in a generally recognized useful product, can one justify the time and resources consumed. Until then it looks like self-indulgence.
- [Cueball sits at a table, eating a meal.]
- Cueball: Can you pass the salt?
- [Cueball pauses, a bite of food on his fork, silently.]
- [Cueball still has fork in mid-air.]
- Cueball: I said-
- Off-screen person: I know! I'm developing a system to pass you arbitrary condiments.
- Cueball: It's been 20 minutes!
- Off-screen person: It'll save time in the long run!
add a comment! ⋅ refresh comments!