Sitemap

About intelligence

Mental processes, scientific disciplines, definition and measurement

--

by Tamas Dancsi

I studied computer science but was always interested in psychology. My brother knew that and was so kind to get me a copy of “Ways of Thinking” by Laszlo Mero, a Hungarian mathematician, and psychologist. In the last seven years, I started rereading it four times, but I’ve only made it through last year for the first time. It was a tough read, like “Hyperspace — A Scientific Odyssey” by Michio Kaku, maybe I’ll tell you more about that in a couple of years as well.

What’s intelligence? How can one measure it? How do humans think and how could computers copy the way they do? What is cognitive psychology, and artificial intelligence? These and similar questions are answered in Mero’s book, I tried to collect the essence of it in this article.

Press enter or click to view image in full size

Mental processes

To understand what cognitive psychology, artificial intelligence, and cognitive science are, we need to become familiar with some definitions first. Mental processes are all sorts of things that our mind can do, they are the building blocks of human thinking.

Intelligence is the intellectual power of using these mental processes to remember and learn, to sense and communicate, to solve new problems and express ourselves. It eventually enables us to experience and think.

We as humans are most interested in learning about human intelligence, but most of the above processes apply to non-human beings, animals, plants and artificial entities as well.

Attention

Concentrating on a discrete aspect of information, while ignoring other perceivable information. A more accurate definition would be “a state of focused awareness on a subset of the available perceptual information”.

Perception

It describes the active process of selecting, organizing, and interpreting the information brought to the brain by our physical senses.

Press enter or click to view image in full size

Memory

Memory is the ability to encode, store and recall information. We build patterns in the short-term (or working) memory and if they seem to be useful and reusable they can make it to the long-term memory. If you tell people a couple of words or numbers, most probably they will remember 7 (±2) of them. The same will work with colors, voices or smells as well. Interestingly changing the time slightly between two stimulus won’t change the result significantly, but waiting longer between doing the test and evaluating the results will. Our scheme-set is constantly changing and the longer you wait, the more new information will start modifying it.

Thinking or cognition

Thinking is a mental action to generate new knowledge by reusing existing information. It’s a cognitive process of understanding through thought, experience, and the senses.

Language

Communication with language is for sure the most sophisticated cognitive skill. Cognitive psychologists may study language acquisition, formation, how language use is involved in mood and so on.

Problem-solving

Discovering, analyzing and solving a given problem to overcome obstacles and find a solution that best resolves the issue.

Creativity

The phenomenon of creating something new and valuable involving the discovery of new ideas or concepts, or new associations of the existing ideas or concepts.

Scientific disciplines

Artificial intelligence

On the one side, it’s a highly technical ability to demonstrate intelligence, learning and problem-solving. Its goal is to construct models for mental processes. For now, the best tool to achieve that is the computer and software.

Get Primed’s stories in your inbox

Join Medium for free to get updates from this writer.

On the other side, one could think of it as a new tool of scientific cognition — in case it’s capable of modeling our knowledge about the mechanism of human thinking. The goal of the functionalities here would be to be as similar to humans as possible, both the results and errors. For example let’s say we built two chess programs, ‘A’ and ‘B’ with the goal of modeling human thinking. Let’s say ‘B’ keeps winning against ‘A’. ‘A’ could still be higher evaluated than ‘B’ in case it models human thinking more realistically.

Press enter or click to view image in full size

Cognitive psychology

Cognitive psychology is a classical field of psychology which studies mental processes. It tries to answer questions like “What is memory and learning?”, “What is remembering and forgetting, association, creativity and all in all, intelligence?”.

It deals with the subject of how humans handle perceived information about the world and what kind of tools are available to use our knowledge to solve new problems. Cognitive psychology aims to present these with laboratory experiments and describe them with exact models and mathematical objects.

Cognitive Science

At the first glance, the field of cognitive psychology and artificial intelligence seems to be similar. They both build computer programs, but for different purposes. Cognitive psychology wants to describe already existing functionalities with models, while artificial intelligence wants to construct similar ones.

This difference led to the birth of a whole new discipline: dealing with abstract objects and trying to model human perception and thinking. Its evaluation system comes from cognitive psychology, but the way it builds its models comes from artificial intelligence.

It states that every aspect of human thinking can be described with suitable computer models. That’s how it differs from cognitive psychology. The latter only builds models for experimenting and exploring, scientists in the field only try to understand more about mental processes.

Artificial intelligence doesn’t guarantee its goals being realizable in the first place. Its fundamental tasks are to build and reuse systems and procedures found to be useful for understanding human thinking. For cognitive science, the model being functional is more valuable than the actual way of implementation, which is the opposite of artificial intelligence.

IQ tests

Intelligence is hard to define. What would you do if someone asked you to sort a group of your acquaintances by intelligence? You wouldn’t necessarily like the idea, but I’m sure you could finish the task. This means you do have a definition of intelligence. We all have, even though there’s still no exact definition.

Psychologists got the task to solve this mystery. They asked different groups of people to sort themselves just like I’ve asked you above. Surprisingly, the results were very unified, test members seemed to have a similar definition in mind.

In the second step, they collected plenty of tests, lingual, mathematical, memory, and association related, logical ones, cultural questions and so on. They reduced those thousands of questions to a couple dozen, that would represent the groups’ results.

Edwin G. Boring’s article, Intelligence as the Tests Test It (1923) lead to the definition: ”Intelligence is what the tests test”. He pointed out that intelligence can only be measured together with special abilities, like arithmetical ability or logical relations. Defining which exact abilities need to be tested are fully culture-dependent. When the separate tests of those special abilities are combined to a total score, the relevance of the abilities cancel out, leaving that total to represent pure intelligence.

At the end it’s the intelligence test, that defines and measures intelligence. Unfortunately, this is quite a unique method, it doesn’t work for creativity or talent for example. There are culture-independent tests as well, but they seem to measure something else.

Press enter or click to view image in full size

Depth of intelligence

Since intelligence is measurable, the question arises, what’s the complexity of it in general. We are going to focus on the results of the standard IQ tests here. To find out an exact number, we need two things, retest reliability and high and low boundaries.

Retest reliability means, that with 75% chance an average candidate would reproduce a measurement twice. We all have good or bad days and are able to concentrate differently on different occasions. Studies have shown quite accurately, that this number is around 5 IQ points.

For standard IQ tests, the range of measurement is between 70 and 135 IQ points, and the results follow Gauss’s normal distribution. 2.4% of the population doesn’t reach the low end of 70 and only 1% is above the high end of 135.

Based on the above we could say that the depth of general human intelligence is about 14–15 skill level difference, which is about as complex as the Chinese board game, Go. By interpreting depth in a highly abstract way, one could state, that the depth of intelligence is similar to the depth of Go.

--

--

Primed
Primed

Written by Primed

Primed is a millennial-led startup company, based in Vienna Austria.