📜 ⬆️ ⬇️

Marvin Minsky "The Emotion Machine": Chapter 1 "Answers to Questions"

The father of artificial intelligence is thinking about how to make a machine that would be proud of us . Marvin Minsky was a rather tough scientist and the fact that he explored the theme of feelings and emotions with his “scalpel of knowledge”, which makes us human, is quite interesting and useful. The book is an excellent example of how to use the “ITish approach” to try to comprehend the “human”: values, ideals, love, pain, common sense.

image

§1-8 Questions


What are character and mood?

We use a lot of words that can vaguely describe what we feel and how we behave. We know that evil people react faster (but usually less carefully) and happy people are less likely to get into a skirmish - but such terms do not give us any idea about how these moods affect our way of thinking. We can look at this concept with the following example: imagine that your car does not start, but when you ask for help from your mechanic, you only get this answer:
“It seems your car does not want to start. Perhaps this is all because of irritation to you, because you treated him badly. ”
But psychological terms like these do not help you at all to acquire any ideas on how to explain the behavior of your car. Maybe you just took too heavy a load in tow and broke some gears of some kind of gear. Or maybe you left the lights on all night, and they completely discharged the battery. In this case, these “mental” terms will not help you in diagnosing and correcting a problem — you just need to know about the parts that make up the machine.

In a situation like this, using the model of the mind as a Cloud of Resources is much better than the model of “One-I” - the new model encourages us to look at parts instead of the whole. Maybe something is wrong with the starter? Is the fuel tank empty? Ordinary psychological words are really useful in everyday social life, but in order to understand how our brain works, we need more ideas about what internal structure it has.
')
To what extent are our emotions innate?

It would seem that normal people share a common set of emotions, such as anger, fear, sadness, joy, disgust and surprise, and some more will include curiosity in this list. However, psychologists cannot agree on whether emotions are innate or acquired. For example, some psychologists attribute anger to the manifestation of fear. This book will not be involved in such debates, because it is more connected with what emotions really are - in the sense of “ways of thinking”, rather than finding answers to their classification.

How does chemistry affect our brain?

Psychologist: Your ideas about switching resources sound good, but can you explain the switching of mental states in this way? Are we not affected by chemicals like hormones, endorphins or neurotransmitters?

There is no doubt that chemicals affect the internal state of our mind - but the widespread belief that they affect our brain directly - is a big mistake - this error is similar to the one that assumes that umbrellas open up due to rain. Here is how one author shows this common mistake:

Suzanne Kaizen: “Too much acetylcholine, lack of serotonin and as a result you will fall into depression. So what happens to the mind? There is a very long path that runs from a lack of serotonin to the thought that the world is “disgusted, flat and useless”. And this way is much longer than writing a play about a person subject to this thought. ”

Also, as the meaning of each individual word depends on the sentence in which it is mentioned, the effect of each chemical substance on the brain depends on how each individual brain cell reacts to it, and each type of cell reacts differently. Thus, the effect of each chemical depends on the reaction of the cells to it, and how the cells that are associated with the cell that reacted to chem. substance, and so on, following a cascade of reactions. Thus, the effect of a chemical substance depends not only on where that substance worked, but also on how parts of your brain are connected. We will discuss this issue in more detail in chapter §§ Chemicals.

How can cars understand what things mean?

In the popular vision, machines do things without understanding what their actions mean. But what does the word "understanding" mean? Even our best philosophers cannot explain what is on the word "understanding."

However, we should not be unhappy with this fact, because this is exactly what should happen with the concept of “understanding”! For most common words in psychology have the same property: the more clearly you try to designate them, the less you understand what it is about. And this property in particular applies to words like “understand” and “mean”!

If you "understand" something in only one way, then we can say that you do not understand anything. For if something goes wrong and you cannot explain something, you will have no room for maneuver. But if you present something differently, then even if a representation is not correct, you can switch to a different view of the object - until you find a way that is best suited for explanation.

The same thing happens when you encounter a new type of problem:

If you know only one problem-solving technique, then you will be at a dead end if this technique does not work. But if you have a lot of techniques to solve this problem, even if you get stuck using one, you can switch to another.

We switch between ways of thinking so quickly that we barely notice that we do it, except when the interaction cascades are so large that changes in the style of thinking affect our emotional state. One of the main goals of this book is to describe the diversity of our mental resources and how they can be organized, and the final chapters of the book will show that much of our human ingenuity stems from the ability to find various ways to solve a problem.

Why do we think that I am inside us?

Philistine: If my mental resources change so much that it makes me feel like I am myself - no matter how angry or happy I am?

We have no evidence that babies are born with the presence of any of these feelings, and, moreover, we cannot trust our childhood memories. Then why do we think that somewhere deep inside the heart of the mind there is an entity that experiences all our feelings and thoughts? As I see it, this could lead to the following:

At the beginning of life, our low-level processes solve many small problems, without any understanding of what mechanisms are involved in solving the problem. And when these processes encounter difficulties, they simply cease to be performed, and the mind begins to do some other action.

However, as we develop several levels of thought processes, these higher levels are trying to figure out "what went wrong" and improve our skills in solving such problems, so we are starting to create new ways of displaying aspects of our recent thoughts. Ultimately, they turn into simplified “models” of ourselves.

Perhaps the simplest and most common model is as follows:

image

However, every normal person also builds many other models that try to describe how we envision such things as social relationships, physical skills, political and economic attitudes, spiritual and social relationships. In chapter §9 it will be described that what each person calls his “I” is a network of “mental models” in which each of the models tries to describe only certain aspects of the human mind.

Why do we have several models of "I"?

Physicist: Why not just merge all these models into a single model that embodies all the features of individual models?

Perhaps it will not be practical, because such a structure will be too large for us to "keep it in mind" as a whole. This suggests that the limitations of our brain should limit us at any given time, as it can only use only a few incomplete models of "I", imposing them on each other, and making up the resulting model - but too incomplete to answer the most questions about yourself.

In addition, for each specific problem, the use of a specific model may be more suitable than the use of other models - due to the set of the most relevant functions for resolving a particular problem. This means that you just need to have several types of models, and the ability to quickly switch between them. Let's hear Richard Feynman:
“... Psychologically, we must keep all the available theories in our heads, and every theoretical physicist who knows physics well knows six or seven different theoretical concepts for the same physical effects. He also knows that all theories are equivalent, and that no one will ever decide which of these theories is absolutely true for a particular situation. But the physicist is forced to keep all the theories in his head, hoping that they will be able to give him a lot of different ideas for guessing the solution of the problem ”.
The key word in this text is “guessing,” because every existing theory has both positive and negative sides; none of the models we have can not fully describe the solution for various tasks or goals, and each of the available models will lead you to a dead end when considering a certain type of problems.

How do we develop new goals and ideas?

The next few chapters will give a general understanding of what goals are and focus on how we acquire them. However, these ideas will be incomplete until we discuss deeper (in chapter § 6) in more detail how ideas work.

In the usual view of how the human mind develops, each child begins with instinctive reactions, but then goes through the stages of mental growth, which impose on the instincts additional layers and levels of abstraction. Over time, the old instincts can still be maintained, but the new layers will gain more and more control - this will happen to such an extent that we can eventually begin to think about our own motives and goals and maybe we can change or present them differently. .

But what base can we use to train self-assessment? How do we choose which goals to adopt and how we justify them? No infant can be wise enough to make such a choice. So, the next chapter will argue that our brain is likely to have evolved in such a way that it copies the ideals and attitudes of our parents, friends and acquaintances!

For the translation, thanks to Stanislav Sukhanitsky.

Table of Contents of The Emotion Machine

about the author


image

Marvin Lee Minsky (Eng. Marvin Lee Minsky; August 9, 1927 - January 24, 2016) - American scientist in the field of artificial intelligence, co-founder of the Laboratory of artificial intelligence at the Massachusetts Institute of Technology. [ Wikipedia ]

Interesting Facts:

Source: https://habr.com/ru/post/354594/


All Articles