A Few More Thoughts about Altruistic Value Calculations

I introduced altruistic value calculations in earlier posts and comments. My intention was to encourage scoring of the altruistic value of more of an individual’s actions, including the means they use to achieve common ends (such as making money).

In one comment I mentioned a numeric 2 dimensional scoring system for relative and subjective score, scale, and distance altruistic value calculations. The first score measured the combined positive altruistic value of consequences of an action. The second score measured the combined negative altruistic value of consequences of an action. This helped me state my intuitions about an action and its consequences mathematically. It also helped me convey how I would score, scale, and compare the altruistic value of two actions in a couple different scenarios of their consequences.

I did not mean to limit anyone’s choice of scoring, scaling, or distance approach.

You can use alternatives as you develop them. For example:
* a single axis numeric scoring system for consequences.
* a binary system of consequence feature counts as scores, consequence counts for scaling, and hamming distances for comparisons.
* a stratified scoring system allowing action comparison within but not across strata.
* a per-consequence system with line estimates for consequence score clouds and axis-specific integration of lines to estimate combined consequences.
* isolation of scoring subspaces for practical use within the total possible scoring space.

I think a single axis integer system would be the intuitive default for most people. People rank individual actions somewhere between bad and good for others overall, without separate of bad consequences from good consequences. In my view, with good information about consequences, a single axis system has utility in many cases. However, altruistic and anti-altruistic values do not cancel each other.

Long-termism and Altruistic Value

The altruistic value of an action with only consequences that change the well-being of one or more hypothetical future entities is null. Put differently, the altruistic value of an action with only hypothetical opportunity costs for a non-existent entity is null. By “hypothetical future entity” I mean an entity that could come into being by undetermined circumstances independent of the action in question but that does not currently exist. By “hypothetical opportunity cost” I mean a cost associated with the absence of existence of an entity for whom an opportunity would otherwise be present.

An altruistic value of null has no score and can not be scaled or used in a distance calculation.

Score

An action that is known to be a contributing factor toward a consequence is not known to be a cause of that consequence. However, assessments of altruistic value of actions are subjective while actual consequences are circumstance-bound and uncertain. Contributing factors are subject to your prediction about their causal nature on a case-by-case basis. Sufficient actions are causal. Necessary actions are only causal if you believe that their consequences will occur.

Scale

Score altruistic value on a numeric scale with a maximum. As you discover actions that you can perform and that cause greater harm or good, reset numeric estimates of earlier submaximal (anti-)altruistic values downward in order to maintain their relative scale wrt your new maximums. An analogous course follows if the consequences of your feasible actions have a smaller maximum (anti-)altruistic value than you thought previously.

Do not use a hypothetical maximum for an altruistic value axis. If you cannot feasibly perform an action with high (anti-)altruistic value, it’s consequences do not define the maximum of your scale.

Allow room on your axes for unintended consequences with unexpectedly high (anti-)altruistic value.

Distance

Distance calculations allow comparison of actions by their relative altruistic values. Accordingly, have several distance calculations and contextualize them. Examples of distance calculations include:
* distance from origin
* Angle distance between actions
* Cosine distance from an axis
* Euclidian distance between actions
* Euclidian distance from an axis maximum

Option Listing and Selection

When comparing actions for their altruistic value, no one necessarily knows how comprehensive or honest their options list may be. The choice of what actions to compare is yours.

Furthermore, having a sense of what course of action is altruistic does not force you to follow that course.

Finally, if you believe, as I do, that people regularly decide against having altruistic consequences despite their own good intentions, you will disavow the common belief that good intentions characterize an altruistic person. A kind heart doesn’t make you altruistic, per se.