Sunday, March 16, 2008

Utility

Using the concept of utility is pretty familiar from various economical and game theoretical problems, at least. But if the framework of viewing the world from the perspective of utility and utility functions is applied to normal interpersonal scenarios, this might be a slippery slope and there are very good reasons for not necessarily voicing the framework outloud. It can often be very much misinterpreted and it can seem like cold and calculating, nothing that people in general like to see in any sorts of relationships, albeit it is very much visible in some of the more utility-focused circles (e.g. the benefit of being friends with someone, etc.). And you seriously don't want a friend of yours or your love telling you that they're with you "because their utility function is at a local maximum when they are around you". That would be quite unheard of.

Yet, even still the utility view allows some room to look at interesting things in relationships between people. If the assumption is that people optimize their utility function, the utility function can be seen as having different time horizons. You can optimize your utility function for a very short time horizon, although this will undoubtedly result in very opportunistic behaviour on behalf of the person. This is of course based on the fact that a utility function that's being optimized on a short horizon will not take into account that many of the long-term implications of short-term opportunism. Then again, long-term optimization may in turn cause the utility function to disregard large moves in the short-term as they even out in the long-term. Analogically this view could be seen in a person that takes shit that is poured on them in the short-term because they see much value in the relationship in the long run. And I guess in theory the latter approach is a good and honest approach, but it too easily allows short-term optimizers to take advantage of you. So the problem is finding an equilibrium in terms of the horizon being optimized. To make the model slightly more complicated, one could argue that the time horizon being optimized should also be variable and in the beginning of a relationship (be it friendship, romance, or whatever) the horizon should be short to mid and then as time passes and rapport is built between the sides the horizon should grow longer. Applied to the real world this would basically mean that you're cautious with new people but give much more leeway once you start trusting someone.

Another thing to consider in utility functions is the internal structure of it. The time horizons do you little good if your utility function is distorted and contains no traces of altruism or other such values. Meaning that if you're an asshole, it doesn't really matter if you're a short-term or long-term asshole as you remain one regardless. This might be another reason for you to keep a variable time horizon against which you optimize. Some people indeed have distorted utility functions where they optimize only personal gain and benchmark it solely against money or position or power. These would quite possibly be the hardball players and sociopaths that progress at lightning speed in corporate environments and the types of people that, for instance, Mintzberg's Power School of Management suggests that are crucial to the success of a company. Of course having a personal relationship with such a person is tedious at best as they will continuously kick you in the head and care little for you even though they may apologize and seem to be sorry.

I guess there are many other dimensions through which one might view personal utility functions, but those were two of the easier lenses. They actually remain fairly well in line with the general understanding that people should have healthy values and avoid opportunism because of moral and ethics. But it will also provide a mechanism for attempting to determine whether or not someone is playing hardball with you, in which case you will have to be able to adapt your utility function for that specific case to be able to survive it. The concept of utility is very interesting, but again, as someone online suggested, there are good reasons why these sorts of things aren't necessarily to be discussed very openly.

No comments: