Continuing from the previous post, I’d like to venture from a discussion of what is, into a discussion of what ought to be.
This post will focus on the following question:
Given an arbitrary value matrix, what actions can be taken to most successfully maximize future utility?
Here, future utility is defined as the time weighted sum of all predicted future emotional states. This question is relevant because the fundamental goal of all humans is the maximization of future utility, so we have interest in knowing how to best achieve our goal.
Since this question makes no assumptions about the contents of the given value matrix, specific proscriptions, like being more kind or focusing on relationships aren’t relevant to this discussion.
For brevity’s sake, we’ll now refer to specific components within a value matrix as values.
To begin, let’s analyze a scenario. Imagine an individual who’s only value is the “desire for good tasting food” and where the activation strength of this value is positive when food tastes good, and negative when food tastes bad. The better the food tastes, the stronger the positive activation, and vice versa for negative activations.
This individual is presented with two plates, one holding a slice of rich chocolate cake. The other holds a glass of ground wheat grass. The individual is only allowed to consume the contents of one of these plates. The individual, being familiar with the taste of both, predicts that their future utility will be higher if they eat the cake, since they predict that the taste of the cake would be better than the wheat grass. The individual then performs the action required to cause this increase in predicted future utility, which is to eat the cake. Upon eating the cake, the individual feels a strong positive activation of their “desire for good tasting food” value, resulting in a strong positive current emotional state.
Let’s imagine the same scenario, but with a slight modification. The individual now has two values in their value matrix: the original “desire for good tasting food” and a new value, “the desire for healthy food”.
The individual is presented with the same two plates. However, in the new scenario, the predicted future utility is different. For the sake of example, let’s say that the individual predicts a net positive future utility when eating the cake, and a net negative future utility when consuming the wheat grass. This is the same as earlier, however, the absolute value of each predicted future utility is now smaller. The predicted future utility of eating the cake is less strongly positive because, although the “desire for good tasting food” value has the same strong positive predicted activation, the new “desire for healthy food” value has a moderate negative predicted activation. Therefore, the predicted future emotional state (the sum of the two value activation strengths) is now only weakly positive, unlike the strongly positive emotional state from before. This results in an only weakly positive predicted future utility.
When taking a given action results in two values whose activation strengths have different directions, we say that those values are conflicting.
The important effect is that the presence of conflicting values results in weaker current emotional states.
Putting this in more relatable terms, we can think of conflicting values as a sense of internal conflict. In this framework, internal conflict robs us of the greater happiness we would be feeling if our values didn’t contradict themselves.
In the above example, the strength of the positive emotional state after eating the cake is suppressed due to the negative influence of the “desire for healthy food” value. The difference in activation strengths between the scenarios with and without the “desire for healthy food” value can be thought of as the negative impact of having conflicting values.
The key takeaway is as follows: If we wish to more successfully maximize future utility, we must minimize the negative impact of conflicting values.
This leads to the central prescription of this post.
Value deconfliction is idea that, by taking active steps to minimize the incidence of value conflict, we can increase the positivity of our average emotional state.
Individual Value Deconfliction
On an individual level, a great deal of the unhappiness one faces is due to the difficultly inherent in balancing conflicting values. How does one find balance between career and family or between education and social life? In this context, value deconfliction has one prescription.
That is, to modify your environment such that the probability of value conflict is minimized.
The process by which humans decide future actions was described earlier as a maximization of predicted utility. The prescription is this: when determining the future utility of an action, place a high weight on the whether that action will lead to more or less value conflict in the future. In other words, do your best to avoid value conflict in the first place.
For instance, an individual choosing between a new high paying job and the strain it will put on their family should have made extra considerations when choosing their career path. Perhaps it would have been possible to choose an alternative occupation that would have had less potential for introducing a future value conflict between money and family.
In order to avoid future value conflicts, an individual must make conscious considerations about the potential impacts of their decisions far in advance of the impact themselves.
These types of considerations aren’t natural tendencies for most, but are likely promotable through practice and cultural reinforcement. The specifics of realizing these techniques on an individual scale are beyond the scope of this post.
Collective Value Deconfliction
Perhaps the most important impacts of value deconfliction are its consequences when applied on a collective scale. In matters of governance, we often see value conflict among elected officials. Officials we elect to promote abstract values like “liberty” or “economic prosperity” can often be seen taking actions directly antithetical to these values. Oftentimes, this disparity is caused by adverse monetary incentives, many times through the action of lobbyists. Here, a lobbyist can compel a legislator to take an action antithetical to a given abstract value by appealing to the legislator’s “desire for wealth” by offering gifts in exchange for specific legislation, or the legislator’s “desire for reelection” by offering contributions to the legislator’s reelection campaign in exchange for specific legislation. Here, the values of the individuals controlling the system are not aligned with the values the collective system is intended to uphold. This misalignment occurs because of the tendency of the controlling individual (the legislator) to experience value conflict between his elected responsibilities and monetary desires. In this case, value deconfliction can guide the construction of improved collective systems that more effectively align individual values with the intended values of the collective system. We won’t explore specific collective prescriptions in this post, but future posts will explore ideas related to this, especially as it relates to charity, governance, and corporate structure.
Benefits and Restrictions
One advantage of value deconfliction is its inherent universality. Because no assumptions are made about the contents of the given value matrix, the prescriptions in this post represent effective means to improve individual emotional states regardless of the cultural background of the individual implementing value deconfliction.
The intractability of considering both all possible sources of value conflict and all methods of minimizing value conflict place natural limits on the effectivity of these prescriptions. However, in order to realize benefits from deconfliction, we need not implement it perfectly, but only to incorporate it as an additional consideration in our normal process of estimating the future benefit of possible actions.