top of page
Search
outml99

PP13 The logic of decision-making



G'day my name is Bruce Robertson and this is Pirate philosophy. In this series of videos, I will be describing an original philosophy, one that you won't find anywhere else but it is one that is logical, rigorous and dynamic. Welcome.


So far in this series of videos we have discussed how making just a few assumptions about a logical processor and its evolution can give rise to many different aspects of thought and life and even self awareness and perhaps consciousness, as described in the previous video.


Today, I want to talk about the process of making decisions. This is a follow up to PP 11 Foundations of Reality and Purpose. Decision-making is what the brain does; it makes decisions so that it can conduct actions, which allow it to interact with the world and to look after its physical body. In order to do this, it creates a model of the world from its sense-data. It has motor outputs, which enable it to interact with the world and it also has internal sense-data, which give it feedback on the physical state of its body and the internal senses, from the body (of well-being or not) come from all the organs of the body; from the skin to the heart, stomach and even the brain itself is an organ and its well-being is also incorporated into the overall well-being of the body. This experience of well-being by the brain, can be labeled as happiness and a lack of well-being as unhappiness.


So from hereon our we'll be using the word 'happiness' to indicate the goal to which the logical processor of the brain aspires. And this goal incorporates not only short-term experiences of pleasure but also long-term experiences of happiness. And the long-term experience of happiness can actually be experienced in the short-term as a feeling of security, that one's long-term goals can be met. As for example, a squirrel can experience contentment or happiness at the start of winter, knowing that it has a good supply of hidden nuts to keep it well provided with food over the cold winter months.


It should be noted that what makes a person happy can be specific to that particular person. The requirements of the generalities of food, warmth, shelter, etc. may be universal, but the specifics of what a person's goal of happiness entails will be unique to that particular person and will be determined by a combination of their genetics, their upbringing and their environment. So, for example, some people are more social than others and want to interact co-cooperatively with others, whereas some other people are more solitary.


So how would a logical processor with a good model of the world go about making decisions with the goal of maximising its happiness? Well, first of all, it would assess the situation. It would collate all the possible options for a decision and then examine those options to determine which one would be most likely to maximise its happiness; and it would do this by extrapolating aspects of its model of the world to indicate the likely consequences of a particular action and then to estimate the likely happiness or unhappiness of those consequences and then it could arrive at an expected happiness of a consequence by multiplying an estimate of the probability of that particular consequence occurring by a parameter, say from negative 10 to plus 10 to indicate the expected happiness; with minus 10 being much unhappiness and plus 10 being a lot of happiness. These values can be exceeded, if required, to indicate, for example, a complete disaster. And those values of happiness might also take into account the time over which the happiness or unhappiness might occur; depending upon whether it is going to be brief, or whether it will last a long time. And obviously, greater weight can be given to those that last longer. I have drawn up a simple algorithm to indicate how this might work in detail. This algorithm is in a generalised form and can be used for all decisions from 'what socks to put on in the morning', assuming one does have socks; to whom one wants to marry or not, as the case may be. I will put a copy of this algorithm in the description below.


Decision Making Algorithm


1. Assemble all possible actions for this decision

2. Select one possible action

3. Assemble all possible consequences of this action

4. Select one possible consequence.

Estimate the likelihood or probability that the consequence will occur.

Estimate the happiness (or unhappiness) that one might expect from that consequence in both the short term and the long-term. (Use some arbitrary units and put on a scale say from -10 ( extreme unhappiness) to +10 (maximum happiness).

Calculate the product of the probability by the estimated happiness to arrive at an overall expected value for the happiness of that consequence.

If There are more possible consequences loop back to ‘4’, else continue

Add all the Expected happinesses for each possible consequence to arrive at a combined expected happiness for this action and store this value alongside a description of the action

If there are more possible actions, loop back to ‘2’, else continue

From the stored values of expected happiness find the one with the highest value.

Select the action that is associated with this value.

Execute the action.


So let's have a look at the algorithm. First of all, assemble all the possible actions for this decision, select one possible action and then looking at that one particular action, assemble all the possible consequences of that action. Then select one possible consequence, estimate the likelihood or probability that that particular consequence will occur, estimate the happiness or unhappiness that one might expect from that consequence in both the short-term and the long-term. Using some arbitrary units, as discussed, calculate the product of the probability by the estimated happiness to arrive at an overall expected value for the happiness of that particular consequence. If there are more possible consequences, then loop back to 4 and go through that loop, for all the possible consequences of that action. And when you've done that, add all the expected happinesses for each possible consequence, to arrive at a total expected happiness for that particular action; then store this value alongside a description of the action. Then if there are more possible actions, loop back to 2 and select another possible action. Assemble all the possible consequences of the action, etcetera and go around that loop as many times as you need for each consequence; and when you've done that for each possible action, you look at all the stored values of expected happiness for each of those actions and select the one with the highest value and then you execute this action.


So I thought I'd go over a worked example of how this would work in practice. Imagine, if you will, that you are the captain of a pirate ship, somewhere in the ocean far out to sea and you detect that you're heading into a storm; you have a decision to make. What are you going to do? Going through the algorithm, you assemble all the possible options. They might include, 'carry on your course', 'head to the nearest port', 'sail away from the storm'.

Worked example For decision making Algorithm


You are the captain of a pirate ship somewhere in the ocean and you detect that you are heading into a storm.

You have a decision to make:

Option1 Stay on your course

Consequences: (A) Possible damage to ship or sails

(B) Get to destination on time

(C) Complete loss of ship and crew

Option 2 Head to the nearest port

Consequences (A) Port may be unfriendly

(B) Delay in getting to destination

(C) Ship is safe

Option 3 Run away from the storm

Consequences: (A) Ship is safe

(B) Delay in getting to destination

(C) Shortage of supplies and unhappy crew




And the possible consequences of carrying on your course might include possible damage to the ship and sails, get to one's destination on time and the complete loss of ship and crew in the storm. So taking each of those consequences in turn; taking the first one: possible damage to the ship or sails; maybe you'd give a probability of that occurring as 20%; unhappiness caused perhaps negative 8; to give a total for the product of the probability and the expected happiness as negative 1.6. Second option: get to the destination on time, likelihood 30%; happiness, say plus eight; taking the product this comes to plus 2.4. Or the third possibility: complete loss of ship and crew;while it is unlikely maybe 1%; but happiness value we're going way over scale and give it a value of minus 100, because it's a complete disaster.Then the product of the happiness and the likelihood, come to negative 1; and that's all the consequences for that particular action. So then we add those products together and we get a value of negative 0.2. So we store that value with the decision to carry on your course.


Option1 Stay on your course :

Consequences:

(A) Possible damage to ship or sails. P=0.2, H=-8, PxH=-1.6

(B) Get to destination on time. P=0.3, H=+8, PxH=+2.4

(C) Complete loss of ship and crew P=0.01, H=-100, PxH=-1.0

Sum of expected happinesses for staying on course: -1.6 +2.4 -1.0 = -0.2


Then you take the next one: Head to the nearest port. I won't go over the details; you can fill in your own values and possibilities of what might happen; perhaps the port might be unfriendly, or you are going to be late in getting to your destination but the ship will be safe or if you sail away from the storm, the ship may be safe but you'll be delayed and there might be other consequences of the delay such as shortage of supplies or an unhappy crew and that sort of thing. But you can fill in your own values for those probabilities and for the expected happiness to arrive at a product for the expected happiness for that particular decision. And then, having done that, you can compare the totals for each of those possible actions and select the one with the largest value. When I did it, the best decision was to head to the nearest port. But when you do it, you might end up with a different decision.


This is, of course, a somewhat idealised algorithm. In practice, there are two major constraints as to its efficiency: first, the accuracy of one's model of the world; the more accurate the model, the more accurate will be the expected consequences and the probabilities. And secondly, the time it takes for the brain or the logical processor to work through all the different possibilities; and it may be that it is expedient to make a decision rapidly, in order to meet the particular circumstances.


So if you are out at sea on your pirate ship, you've got to make a decision quite quickly. You don't want to delay too long, otherwise, the storm will hit you before you've made a decision. On the other hand, if there's an important decision to be made, where there is no time constraint, it might well be beneficial to allow the subconscious brain to work through all the different possible options and consequences; perhaps even during sleep. Hence the common expression for when facing a difficult decision: 'I'll sleep on it'.


One more note on a technical issue: it is not necessary for the probabilities of possible consequences for a particular decision to add to 100%. This is in part because the broad probabilities are only rough estimates and also because if the estimated happiness of a consequence is zero, or close to zero, i.e. neither happy nor unhappy, it can generally be ignored because the product of the probabilities and the happiness will be zero.


So does this algorithm fit the facts? We have seen how an algorithm of this form follows logically from the assumptions that were made at the start of this journey. But does the algorithm fit the facts as we observe them? It would seem that it does. The decisions and actions of other people are entirely compatible with the idea that they are seeking to maximise their happiness. People may not be aware that this is what they're doing, as the processes of the algorithm may be done in the subconscious, at a level in the brain or pyramid of patterns, that is below the level of consciousness.


It should be noted that this account of the decision-making process is descriptive, and not proscriptive. In other words, I'm not saying that people should make decisions to maximise their happiness, I am saying that people do make decisions that seek to maximise their happiness.It is the logical thing to do. And this applies to all people. Whether this is a beneficial thing for people to be aware of is a moot point but it is certainly useful for understanding the decisions and actions of other people; i.e. that their decisions and actions are motivated by their desire to maximise their own happiness.


Decisions are hugely important, they are what the brain does. And the decisions made and the actions taken have consequences that can last the rest of one's life. And the quality of those decisions taken will influence the happiness one achieves in the rest of one's life. And the quality of decisions will depend upon the accuracy of one's model of the world and also on one's understanding of what will make one happy in the long run and what will not. And so it is important to learn the specifics of what will make one personally happy. It is also important to create an accurate model of the world; and for this it may be necessary to try new and different things and to explore one's environment, both physically and socially. While this may not bring immediate pleasure, it may well improve one's model of the world so that one can make better decisions in the future and then an in this way, improve one's prospects for long-term happiness.

Finally, I would like to note that despite the paramount importance of decision-making, it is a topic that has largely been ignored by Standard Western Philosophy. This is a failing of Standard Western Philosophy and is a major point of difference between Standard Western Philosophy and The Pattern Paradigm philosophy that I am describing.


Well, that is all I have for you today. If you have any interesting comments or questions about today's video, please leave them in the comments section below. And if you would like to continue this journey with me, then please subscribe to my channel, give it a thumbs up and ring the bell.


Thank you.




11 views0 comments

Recent Posts

See All

Comentarios


bottom of page