The original author has put forth a speculative thesis--namely, that the INTJ is reluctant to make social commitments too long in advance and that this is connected with his need to make a choice based on the optimal data set reasonably attainable preceding his decision. Analytically, the question of interest is, if the prototypical INTJ is supposed to be this mastermind of a personality with an immaculate system of reasoning and intuitive faculty for grasping contingencies, how can a person so stolid about planning their goals be so indecisive with respect to people? Perhaps the paradox of an INTJ is best illustrated in the personality of the notable economist John Maynard Keynes. As an example, Keynes noted that in the long run we are all dead. But why would a thinker that holds such a view publish a paper in 1930 called 'The Economic Possibilities for Our Grandchildren'? Obviously, Keynes himself was looking with a view to the future and the conclusion one might draw is that while he masterminded a new school of economic thought he was nevertheless a capricious personality. We know, also, that Immanuel Kant who almost never broke his routine for anyone or anything (absent when he first read Emile) would take a long lunch if he was enjoying the conversation. Accordingly, we see with Keynes and Kant that, whether grandchildren or conversation, the common pattern is they were resolute in their areas of expertise but maintained a socially flexible posture in their respective societies. Why is this the case?
As a first conjecture, as much as economists assign algorithms to human behaviour (such as GDP) and philosophers propound metaphysical theorems, in the final analysis it is impractical to have a teleological account of social behaviour. We know, for instance, that there are more possible chess games than elementary particles in the universe, and more logical possibilities of reasons and choices than chess games. As an example, if A has to make a decision to B to perform C at some time in the future, but just before C there is another event X that will in effect put A in a different mental state than the time of making the decision (and possibly for the worse), then it would follow that there is risk of lost utility. If humans were invariant with respect to all stimulus, then the question of X would be immaterial to the performance of C, but most humans are not this consistent. From a Benthamite point of view, this sort of calculation would seem plausible from the perspective of utility maximisation. Hence, also, we can deduce that, whether one is this person or encounters another with reservations as such, one can reasonably conclude that they are weighing their options in the pursuit of rational utility maximisation. But we need to break this idea down still more. At loggerheads is the freedom to socially contract with other humans (something no doubt fundamental as the desire to travel to see other humans for social intercourse and recreation is fundamental to being human) but also the desire not to be bound by obligations prematurely should something better arise. That something better should possibly arise follows from the evolutionary rather than teleological nature of socialisation and thus a decision that appears competitive at t1 may not be the fiercest choice given the conditions at t2. Now, the future is not an entirely unknown quantity; the conditions of tomorrow follow from the conditions of today. Accordingly, the rule of thumb is to not make plans too far in advance given the fickle nature of people (although when everyone does this it collectively creates fickleness, but inconsistent people are not entirely unreliable so long as they are consistently inconsistent) and also because if one does one might miss a lot of the big things. Imagine that, just as a conversation became lively and intensely philosophical, Kant had to remark that the hand as just struck x and it time for his company to go home. The opportunity cost of knowledge foregone would be astronomical!