By an 'agent,' I mean a conscious being with a body that interacts with the physical universe. For our purposes, we will consider time as not branching, linear and infinitely subdivisible.

The agent has a body, which exists within time and space. It has a stream of conscious experience associated with it. The body is seen to perform actions. There is some relation between what is experienced, and what the body does. If a desire is experienced, then the body acting in such a manner to satisfy the desire may follow. It would be good to have a classification of all subjective mental states, to say when each occur in terms of the physical state of one's brain and body. It would be also good to have a description of how people act, and how that relates to conscious experience. We could start by putting a coarse structure on it, and then trying to continually refine this structure.

Ideally, we would put portable MRI scanner-helmets on our heads and wear them all day, but unfortunately no such thing exists. We would also put video cameras in every room in our houses and everywhere else we went, and film all of our actions throughout the day. Afterwards we could analyse the relations between what we had done, what we remembered feeling at the time, and what the brain scanner had recorded our brain as being at the time.

The agent in time

The agent is seeking goals most of the time.

Structure of an agent and its relation to the rest of the world


In creating an artificial intelligence, the difficult bits are the mind and the interpreter. It's questionable whether artificial intelligences would have a soul.

  • The interpreter takes sensory data and places a structure on it. It tries to jam this data into pre-existing theories. Vision, balance and sight are used to create an internal model of a spatial environment. Alphabetic writing is interpreted to be a sequence of letters, divided into words. We see that learning can affect how sensory data is interpreted. The ability to change your interpreter (for example to process symbols) may be part of what makes humans different from lower life forms.
  • The soul is subjective experience, which depends one's knowledge of the world at the present. The arrow goes back to the interpreter because 'you know what you are feeling.' The origin of the emotions and sensations is semi-subconscious. You may feel a sense of friendship, but not have in your conscious mind an understanding of why you feel this way and why it is good that you do.
  • The mind is a process of breaking apart and recombining theories. It doesn't matter what algorithm is used to do this as long as it is fast enough. You could have different kinds of mind that in the end behave in similar ways. All that matters is that the 'threshold of intelligence' is passed. An arrow goes back to the interpreter because 'you know what you are thinking' (although it may be largely subconscious). Well-fitting theories get kept. In human beings the decision on which theories to be kept (transferred from short-term to long-term memory) may occur when sleeping, when the mind isn't doing anything else.
    • This process of 'keeping theories' is not merely an aggregation of distinct theories, but an incorporation with and relation to everything that has hitherto entered into one's memory. It is commonly thought that an artificial intelligent machine, such as an android, would have perfect memory of everything he did and be able to calculate in his mind very quickly, because todáy's computers can store a huge amount of information perfectly and do arithmetic very quickly.1 However, storing information is no use unless it can be retrieved and used. A mass of unstructured information is no use for this. Storing a large amount of information is easy: if we say the information has size n, then this can be done in O(n). The hard part is to structure this information. Knowledge of the world should reflect the world, and when someone is thinking of one subject, they shouldn't have completely unrelated subjects jumping to mind. Learning a new body of knowledge (like learning how to fix car engines or operate heart surgery, or to speak Cantonese) is difficult, because the process of putting a structure on this knowledge could be as bad as O(n2), depending on the exposition. To facilitate learning, expositions of knowledge should mirror an internal representation - such as with 'mind maps.'
  • Habits are processed by the mind in exactly the same way as theories. A habit is a mechanism for doing something. The course of one's actions can be viewed as a sequence of actions, each going along a certain algorithm, for example walking along a known route from place to another.
  • The decider picks a habit to perform based on unchangeable desires (possibly a desire for dopamine release in the brain) and the model of the current situation it gets from the interpreter. The interpretation will have included predictions about the likely outcome of various actions.

Perception and conceptualization

These are formed out of 'fragments' of sense data.

Reasoning, understanding, generalization

People notice common themes. They form general concepts, and likely forget what they were based upon.

I can't think of an example at the moment (which proves the point!). To prove this to yourself, every time you come up with a belief, ask yourself, 'Do you have an example of that? like a particularly smug job interviewer. You are likely to come up with the response, 'No, I can't give you any specific examples at this precise moment, but it is true nonetheless, according to my general experience of things.'

This might seem to be an individual experience of mine, but I believe strongly that it is in fact part of the way that everyone works.

A - What is cheese?
B - Cheese is a type of food made from milk.
A - But how do you know that?
B - From my general experience of hearing the word used and deducing its meaning.
A - On what occasion did you first work out its meaning?
B - I can't remember.
A - Tell me about a time when you deduced the meaning of the word.
B - I can't think of one.
A - What is the earliest occasion that you knew what cheese was?
B - I can't remember.
A - I don't think you know what cheese is at all, which is an essential aptitude for this cheese-making position. Get out of my cheese factory!


Differences in personality from one person to another correspond with with quantitative differences in physical brain structure (quantities of chemicals, proteins, hormones, various neural circuits, etc.) I believe that all human specimens have the same structure - feelings, desires, needs and capabilities and so on - but these parts of the structure can be of different strengths. These differences lead to qualitative differences in action and one's social position.2

Causes of personal characteristics are conventionally divided into 'nature' and 'nurture,' and I have no reason to diverge from from this. For example, someone with an aggressive expression may have gained the perception that other people are out to get them. They may be of low intelligence and be unable to understand the affairs of the world or to find a way to cope. On the other hand, they may have an inbuilt tendency towards this kind of behaviour.

I'd like to have a model of agents which explains different aspects of people's personality.

Goodness = altruism + ability.

Altruism is a very abstract concept, that one looks for a 'good outside the self.' It only manifests itself in specific examples.

  • An older man in a Bronze Age village one sunny autumn evening, teaching a young lad how to weave baskets.

Ability could be intelligence, physical strength, or determination (among other things).

It's also true that how we perceive other people's personalities and behaviours depends on our own state of mind. Consider walking alongside a busy road. Within each car, there are one or more unique individuals. Some may be very happy. Some may be hungry or tired or sad. When we are feeling good, we imagine that everyone else feels the same way, excited about what the future has to bring. When we are feeling down ourselves, everyone else tends to become somewhat threatening.3 But how the people within the cars actually feel does not depend on us, and a whole range of feelings is distributed within them.


Individuals don't just feel, but also act. Their actions are based on a perception of the world, and what would happen with different avenues of action. (A catchy way of putting this is: Your actions show you what you really believe. I'm not not sure if there is an original source for this statement and I have seen it in various places.4)

There are two illustrations of this way of viewing human behaviour. In both cases motivation is something to be deduced rather than directly observed.

One of someone carrying a pile of coloured and patterned masks, with different facial expressions. He lifts one mask for a moment, and then he lowers it and another takes its place. All that we see are the masks, and the true motivations are hidden to us, and to the intellect of the agent themselves.

The other is of the agent as a puppet, with a little demon pulling the strings.5

It would seem best for some avenues to seem attractive, leading to striving to reach the state that they lead to; and others to be feared, leading to striving to avoid that outcome. It wouldn't make sense to feel good about everything, but at least then if you have perceived correctly, whatever the outcome you will be pleased with it. Worse than either possibility would to be pessimistic about every possible outcome, leading to inaction. Emotions are distributed according to contingent circumstances, but not completely so. Someone who has been sentenced to death by firing squad may still feel angry, upset and afraid, even though none of these emotions can motivate any useful course of action, assuming his execution to be inevitable.

However, perhaps it is best that our feelings don't completely adapt themselves. If our prisoner keeps on dreaming of life and freedom, who knows, there may come a chance for him to escape and him take it. Although the situation may seem depressing and hopeless, if we keep on dreaming of a better world, it means we have the chance of seeing some way of changing things and taking the appropriate actions, however out-of-the-ordinary or extreme they may seem.


  • What emotions we feel at any one point in time can be explained to ourselves as a correlate of our subjective experience hitherto. This experience itself, however, contains emotions. For example, if someone has gone through a lot of pain to reach present success, then the feeling of success will be heightened by the former pain. So we have emotions relating to emotions. Is there a 'ground' for emotions - a source for emotions that is not emotions themselves? I need to consider more stories of people's lives and more potential unfoldings to answer this.
    • I think the ground is a few things which relate strongly to survival, like being loved, having things to do, learning new things, having challenges, fear, having a full stomach, and so on. Over this ground further emotions are created — so if you anticipate eating you will actually feel happy, or if you remember a happy occasion, and you might get angry by fantasizing about someone being nasty. (I find I get angry at purely hypothetical situations (as well as real situations, and situations recounted by other people, and situations that were a long time ago, and…))
      • This could be tested - put some rats in a cage where they are subjected to the smell of food, and compare them to rats that haven't. Then either see which set dies first, or is the least lively, however you would do that. I expect that the rats subjected to the smell will be hungrier and more tired, because their bodies have been using up energy (which in a human being feels like being happy and awake) in anticipation of having more soon.
      • Although coffee has a similar effect to food in its immediate effects, there is a striking difference: being about to eat feels good, but being about to drink coffee feels bad. (Consider the stereotype of someone early in the morning, shouting "Durrrrr… Coffee, coffee, where's my coffee!!!!") )See [] for a further demonstration of this point.) That's because it contains no nutrition.
      • There are other questions we might ask to expand on this point. Suppose concept A with ground association $G_1$ at time $t_1$ is invoked in an experience that forms concept B. Percolation therefore gives B a ground association $G_1$. Later on at time $t_2$ experience alters the ground association of concept A to $G_2$. Does concept B retain $G_1$, or does it automatically acquire $G_2$?
        • Concrete example: … something to do with cakes, blowtorches, and coloured ribbons …. finish this later….
  • I think one's life experience is a bit like a tree. The brain is like a massive pattern matching machine. Emotional associations of an experience percolate through from previous experiences, which are all being activated and are in the periphery of one's consciousness. (This means that it is impossible for someone to understand themselves for more reasons than the obvious one ­— that the understanding would have to include itself, but it isn't big enough. Someone could feel uneasy, and not know why. That's because there is a previous experience that has this association. It's the same with feeling angry or feeling resentment — hence the previous bullet point. You don't have the ability to traverse your semantic web at will, and it might actually be a really bad idea for a living creature to have that ability. They would probably go off and explore the contents of their brain, and develop an understanding of it within it. Attention that goes into this is not directed towards other activity that could be more useful, and in understanding itself, the initial understanding might be changed for the worse.) We should aim to list all of the different emotional associations that can be associated with experiences — that would be a great achievement.
  • Attention is drawn by bright lights. You feel that it was a conscious choice to look in a particular direction but there is something deep and ancient going on. An example of this is a television which is on in a room. Another example is a computer monitor. In my current job I have a computer monitor in front of me. When I am working away from the screen, I sometimes turn the brightness down so that my attention is not drawn by it. Another example is at home, turning the lights off in the room. This helps to concentrate on the computer screen.

We want to describe agents in a systemic way. There's a gap in our description, which is the object of study of the humanities.

Unless otherwise stated, the content of this page is licensed under Creative Commons Attribution-NoDerivs 3.0 License