AI and the Meaning of Intelligence — Human & Otherwise
The word intelligence is unhelpful, as it promises insight but leads mostly to confusions about it constituent parts. Both in discussing and understanding AI and people. So we need a new term… How about KP?
The word Intelligence has become a kind of ‘black box’, not unlike AI itself, that contains a ragbag of elements and that has changed over time.
Like ‘Art’, ‘Love’ and ‘Capitalism’ is is a word that means different things to different people and at different times.
Humpty Dumpty said, “When I use a word, it means just what I choose it to mean — neither more nor less.” Alice in Wonderland
This of course tends to obscure rather than elucidate the various abilities both of AI and people.
More like a useful shorthand than a scientific term the concept or IQ, which has shaped and dominated the concept of intelligence has been responsible for much misunderstanding and misclassification of people. Most notably the neurodiverse and other minorities.
Ever since 1904 British psychologist Charles Spearman argued that there is a single, general factor that underlies all intelligent behavior and called this factor g for “general intelligence.” Based on Spearman's observation of correlations between scores on different cognitive tests.
Though widely used, even in psychology, this ‘g’ remains an unproven speculation, now explainable by other known mechanisms. Including the tendency of skills in one area to inform, support and bleed over into others — not to mention the relative wealth, education and social status of test subjects.
A whole book, let alone an article, could be devoted to exploring the issues arising from the dubious uses and misuses of IQ test from inception to recent times in directing people and their life chances, but let’s leave that for another time.
Here’s a brief history of term provided, appropriately by BARD, Google’s AI:
As we see the term was originally used broadly and in an operations sense to denote the ability of a human to understand and act on information.
It ten broadens still further to include “any kind of mental ability, including memory, reasoning, and creativity.”
It’s interesting to note here that this is the point where ‘creativity’ get lumped into the term.
Alfred Binet developed the first IQ test, which is designed to measure intelligence in children, in 1905:
“The first IQ test was developed by Alfred Binet in France in 1905. Binet was asked by the French government to develop a test that could be used to identify children who were at risk of academic failure. He developed a test that consisted of a series of tasks that were designed to measure a child’s reasoning, memory, and problem-solving abilities.”
Which all serves to show not only what a rag-bag of a term (and concept) ‘intelligence’ is and remains, but how unhelpful it is in teasing out an understand of either human or ‘AI’ capabilities — let alone the how.
How either creates and provides those capabilities — and both the capabilities and how they’re generated differ.
To do this we will need to put aside our two sided, love/horror romance with the term AI, and with the term intelligence itself. To borrow an analogy from physical science, which recognises ‘Element, Compounds and Mixture’, intelligence as a concept is clearly a mixture, which we are treating as if it it were more like a (chemical) element.
In fact it seems to me we seem to be stuck at the stage before the four classical elements (fire, earth, air, and water) where propounded by Empedocles, when Greek philosophers debated which substance was the arche (“first principle”), or primordial element from which everything else was made.
We now know that not only are there compound made up of elements but the ‘standard model’ maps what those elements are made of — and on down.
While no one know how far down the ‘Intelligence Rabbit Hole’ will go (although there are some tantalising intimations or entanglement between the quantum level and the role of consciousness and the ‘observer’ only just being explored) it’s easy to see that a continued reliance on such a rag-bag term is only go continue to confuse and hold us back.
Meanwhile is we take the word ‘intelligence’ from AI what might we replace the terminology with, is a valid question!
Here’s my first suggestion: KP. Knowledge Processor.
I suggest this not because of the AI ‘hallucinations’, but because this seems to me the most apt description I can think of of the leap forward which is the current crop of ‘AI’ — LLMs embodied in chatGPT etc.
AI = A KNOWLEDGE processor — significant advance on data (and word) processors
AI, as it stands today, exemplified by ChatGPT and derivatives is probably best termed as a ‘Knowledge Processor’ — in the line of succession from ‘Data Processing’ and (to a limited extent) Word processing. (Which in the context might be better termed character or letter processing).
But this is but a start. Perhaps current KPs can now help us to refocus and better understand what is really down the Intelligence Rabbit Hole?
Copyright Barry E James, April 2023
NOTES
**Timeline of the Origins and Usages of the Word Intelligence**
* **17th century:** The word “intelligence” is first used in English to refer to the ability to understand and act on information.
* **18th century:** The word “intelligence” begins to be used more broadly to refer to any kind of mental ability, including memory, reasoning, and creativity.
* **19th century:** The word “intelligence” is used in the context of psychology to refer to a general mental ability that is responsible for learning, problem solving, and adaptation.
* **Early 20th century:** Alfred Binet develops the first IQ test, which is designed to measure intelligence in children.
* **Mid-20th century:** IQ tests become widely used in schools and other settings to assess intelligence.
* **Late 20th century:** The popularity of IQ tests leads to a greater focus on intelligence in society.
* **21st century:** IQ tests continue to be used, but there is increasing debate about their validity and usefulness.
**The Original Ideals Behind the IQ Test**
The original ideals behind the IQ test were to provide a way to measure intelligence in children and to identify those who needed additional educational support. The test was not designed to be used to rank people or to make decisions about their future.
**How the Popularity of the IQ Concept Has Shaped the Usage of the Word Intelligence**
The popularity of the IQ concept has shaped the usage of the word intelligence in several ways. First, it has led to a greater focus on intelligence in society. Second, it has created a sense that intelligence is a fixed trait that cannot be changed. Third, it has led to the use of IQ tests to make decisions about people’s lives, such as where they should be placed in school or whether they should be hired for a job.
The use of IQ tests has been controversial, and there is a growing body of evidence that suggests that intelligence is not a fixed trait and that it can be improved through education and other interventions. However, the popularity of the IQ concept is likely to continue, and it is important to be aware of the potential biases and limitations of IQ tests.
BARD AI
