Assignment 2

by Seum-Lim GAN (gan@media.mit.edu)

Part A

1.1 The agents of the mind
Everything is NOT mindless. If all things are mindless, how can they do what they are supposed to do ?

Refutation :
All things are mindless. Each and every agent only knows how to do one function and good at performing that task. There is not need knowledge to perform a single very specific task when a set of rules are defined and done. It does not matter how many times it is done, it is the same one job. Each agent is programmed to do it well. Having a program does not mean that it has a mind, it simply means following the rules.

2.1 Components and connections
I'll say there should be interconnections among agents instead of just through the "top" agent".

Refutation :
If there are interconnections among agents, everything will become a mess as there is no control. An agent taking orders from two or more other agents will probably not able to do anything at all since it does not know which order to execute. An agent should only take order from one other agent to perform its task. It should be able to give orders to one or more agents and wait for other tasks to be completed if needed.

3.1 Conflict
There is a thing called "urge" which makes us want to do one thing rather than the other. If our eat agent grew too strong, we can still learn to supress it. That is how "fasting" work.

Refutation :
Nope. This urge will still breakdown evetually and taken over by other urges or stronger agents. If we can supress urges, we will not live a normal live. Those can indeed supress urges would often end up being "mad" or "crazy" and "loose their mind".

4.1 The self
The self could be the whole object or entity and nothing else. Or, it could be parts or groups of the entity and not the whole entity. That's why it is not definite.

Refutation :
Self should be made up of may selves that change state very often. When their states at a given time are known, we can define this whole self to be the self.

5.2 Unaswerable questions
The idea of "cause and effect" is wrong if we need answers to the last question. However, there is no last question as we can never reach it when there is cause and effect around.

Refutation :
If we say "cause and effect" is wrong, then we will not get any answer at any level at all. So, it cannot be wrong. The only time when we say a loop has occured is that when we do not understand the question and the problem. We need an answer anyway, so we will make that question the same as other questions in the higher level which will create a loop.

Part B

6.4 B-Brains criticism :
If the brain is divided into two parts as it is physically (left & right), then when A brain is damaged or dead, the B brain will be of no use since it has no way of knowing what is happening in the outside world. OR, when B brain becomes damaged of dies, A brain will not know how to re-orientate itself (since B brain cannot inhibit it); it will keep repeating itself (since B brain cannot stop it); does anything at all (since B brain cannot tell it if it is good or bad); gets occupied with too much detail (since B brain cannot stop it) and etc..

if this is the case, then we will not have even partial function of our movement or speech for example, when one part of our brain is damaged.

The same applies when there is more level or brain as C brain, D brain and etc.. When something is wrong somewhere along the chain, anything below that will become useless and anything above that will be hopeless.

Proposal
How about the idea of many brains working in parallel ? This way, we could explain why when some brains fail, not every function we have is lost.

Both parallel and serial brains exist and work together as a massive unit. Some parallel brains will need to support or be supported by other serial brains. So, we have a cross-linked network of brains.

Addendum :
In section 11.8 about half-brains, you mentioned that when some major part of the brain is damaged, some mirror region (which I'll say is another brain because it is in a different location) will be able to take over its functions. This explains the parallelism of the brain.

7.8 Difference-engines criticism :
If the brain is an "equilibrium" machine, then whatever we do will not be to any extreme. Since the goal is set, the brain will want to keep as close to that goal as possible.

When the real situation becomes too far off from the initial goal, the brain will stop or will try to do anything at all to bring the situation back to within "bearable limits".

Many people have very high expectation of things before they start doing them and usually will not reach that goals most of the time. If the difference-engine is how the brain works, then why are there so many people who give up when they cannot reach their goals ? Only a handful will continue to struggle until they reach the "bearable limits" or even the actual goal. Some will die because of physical limits but just before they die, they are still determined.

8.10 Levels and classifications criticism :
If we consider things in a logical way, the level and classifications seem to work. However, human beings are most of the time, not logical at all. There are people who would say that a bird is not an animal.

Things could just be arranged in a long list and the brain will search through everything one by one until it finds the right thing and it does this real quick. So, there is no need to put things in levels.

9.4 Enjoying Discomfort :
I don't have any criticism to this idea but rather will have something to add to it. People "enjoy discomfort" because of many surrouding factors like peer pressure (like afraid of being laughed at by friends, especially in kids); trying to hide their fears and being stubborn.

This is like 9.3 learning from failure when after this discomfort, we will try to avoid getting into this position and forcing ourselves to enjoy this discomfort again in future.

10.3 Priorities criticism :
If we place things in priorities when we want to decide them about something, then many a time we give the wrong answer. This is because when assigning the priorities, we may give a higher priority to some agents because we want it to be that way even if it is not. Or, we may not even have an agent to give the right answer because we have not learnt how to create one to do this.

Proposal
We probably get all the agents we have to decide what something should be and then some analysis agent will look at all the answers to see which one should be logical/correct based on known experiences.

Both the original idea and the new proposed idea will work at different times depending on the situation. For example, in a judge duty, we may look at all the answers and select the one we think best. In a biased situation, we will assign priorities to all answers and select the one that benefits us most or makes us happiest (by chance, the correct answer may be chosen).

11.8 Half-brains criticism:
I agree with the part except that the mirror part of the brain performs lower level functions and less administrative functions.

I'll say they should be working together and doing both low and high level functions. When one is damaged, all is not lost, the surviving part will be able to do what it can do and as time goes by, can be trained to do the part of the dead brain.

Back to my Society of Mind main page

Gan's Home Page | Start Page