Hello, dear Readers,
Today, I want to discuss a topic that is personally meaningful to me. With my background in psychology, I'm often bothered by specific trends I observe in the tech discourse. Let me explain.
Over the years, I've noticed (and I’m sure I’m not the only one) that specific terms and portrayals of psychiatric conditions have been misused in pop culture, in public discourse, and sometimes in tech conversations, leading to discrimination and a lack of education on, for instance, mental health conditions. Take bipolar disorder, for example. Many people mistakenly associate it with dissociative personality disorder, which is a distinct condition with its own complexity and challenges. Another misconception revolves around schizophrenia, where people often believe it exclusively pertains to delusional and violent individuals, while reality is quite different. Recognizing this as a widespread issue, I'd like to delve into how psychology and psychiatry terms are misused in the tech discourse.
While some of these misuses may not be as extreme and fearmongering as the examples given, the appropriation of these terms by the tech industry still distorts their original meaning in the public eye. Today, I want to focus on the term "Intelligence," as it lies at the core of our perception and understanding of Artificial Intelligence.
Defining Intelligence
Interestingly, the concept of intelligence has long remained undefined in psychology, and to this day, it lacks a single, universally accepted meaning. Different schools of thought propose various interpretations of what intelligence truly entails and how to evaluate it. Rather than providing you with "the" definition, I'd like to share a couple of examples from my favourite representations of intelligence. These examples help us grasp the expansive range of what intelligence can encompass.
The Cattel-Horn Theory of Intelligence: Factorial Intelligence Model
The Cattell-Horn Theory of Intelligence, known as the Factorial Intelligence Model, provides a comprehensive framework for understanding intelligence. According to this theory, two main types of intellectual abilities are nested under general intelligence. The first is crystallized intelligence (or ability), which refers to accumulating knowledge and skills acquired through learning and experience. This includes vocabulary, general information, and expertise in specific domains. Crystallized intelligence is often measured through tests that assess one's ability to recall and apply acquired knowledge.
The second type is fluid intelligence (or ability), which represents the capacity to solve new and unfamiliar problems, think abstractly, and adapt to novel situations. Fluid intelligence involves mental processes such as reasoning, problem-solving, and pattern recognition. Unlike crystallized intelligence, which draws on previously acquired knowledge, fluid intelligence focuses more on cognitive flexibility and applying logical thinking to new challenges. It plays a crucial role in tasks that require adapting to changing circumstances and generating innovative solutions.
In addition to crystallized and fluid intelligence, the Cattell-Horn Theory later introduced the concept of visual intelligence (or ability). Visual intelligence refers to the mental processes used to handle visual-spatial tasks. This includes skills related to perceiving and manipulating visual information, such as recognizing patterns, mentally rotating objects, and understanding spatial relationships. Visual intelligence plays a significant role in navigation, architecture, and artistic expression tasks.
Multiple Intelligences Theory: Non-Factorial Approach (by Howard Gardner)
The Multiple Intelligences Theory proposed by Howard Gardner suggests a non-factorial approach to understanding intelligence. Gardner proposed that intelligence comprises multiple distinct categories or "intelligence" that are relatively independent of each other. According to his theory, individuals can possess varying degrees of proficiency in different intelligences.
Gardner initially identified seven intelligences: linguistic, musical, bodily-kinesthetic, logical-mathematical, spatial, interpersonal, and intrapersonal. Later, he added naturalist intelligence as another category. Each intelligence represents a different area of human potential and encompasses specific cognitive abilities.
For example, linguistic intelligence involves analyzing information and creating products using oral and written language, such as speeches, books, and memos. Musical intelligence is the capacity to produce, remember, and derive meaning from different sound patterns. Bodily-kinesthetic intelligence refers to using one's body effectively to create products or solve problems, such as in sports or dance. Logical-mathematical intelligence involves the capacity to develop equations, make calculations, and solve abstract problems.
The other intelligences include spatial intelligence, which deals with recognizing and manipulating spatial images; interpersonal intelligence, which involves recognizing and understanding other people's emotions, desires, motivations, and intentions; and intrapersonal intelligence, which refers to self-awareness and the ability to recognize and understand one's own emotions, desires, motivations, and intentions.
Gardner's theory emphasizes that each intelligence can develop independently and autonomously, influenced by biological factors, personal experiences, and cultural contexts. It provides a broader perspective on intelligence, highlighting how individuals can excel and contribute to society beyond traditional measures like IQ.
Understanding the Complexity
These two frameworks, the Cattell-Horn Theory of Intelligence and the Multiple Intelligences Theory demonstrate human intelligence's complexity and multidimensionality.
Further, when discussing concepts like IQ (intelligence quotient), we evaluate a specific facet of intelligence or cognitive abilities. IQ tests primarily measure cognitive skills, such as logical reasoning, problem-solving, and verbal comprehension. Similarly, when we refer to EQ (emotional intelligence), we specifically address a person's ability to recognize, understand, and manage their own emotions, as well as perceive and navigate the emotions of others. These examples highlight that intelligence can be broken down into dimensions, each capturing a specific aspect of human cognitive or emotional abilities.
By understanding and appreciating the different facets of intelligence, we can better appreciate the diverse capabilities of individuals and avoid oversimplifying or misusing psychological terms when discussing topics like artificial intelligence.
Artificial intelligence
When we look at how organizations like the OECD and the EU define AI in the AI Act draft, it becomes clear that intelligence can be understood differently. According to their definitions, an AI system is a machine-based system with specific goals. It uses input to generate outputs like predictions, recommendations, or decisions that can impact the real or virtual world.
It's important to note that while AI aims to mimic certain aspects of human intelligence, it doesn't cover the entire range of human intelligence. Instead, AI focuses on specific cognitive abilities that are relevant to its objectives.
For instance, AI excels at tasks like image recognition, where it can identify objects in photos (think about how your smartphone can recognize faces in pictures). It's also good at natural language processing, which means it can understand and generate human-like text (like when you use voice assistants to ask questions or give commands). Additionally, AI algorithms are great at decision-making, such as providing personalized recommendations (like when streaming platforms suggest shows or movies based on your preferences).
While AI algorithms are excellent at recognizing patterns and processing data, they lack the broader understanding, context, and personal experiences that shape human intelligence. AI systems are designed to achieve specific goals, like personalizing recommendations or automating tasks, and they prioritize efficiency and accuracy in achieving those goals.
Unlike human intelligence, AI doesn't possess the same level of creativity, intuition, empathy, or moral reasoning. These complex cognitive processes are unique to human intelligence and enable us to navigate the world with flexibility and adaptability.
By understanding these differences, we can appreciate both the distinct qualities of human intelligence and the specific intelligence exhibited by AI systems.
Conclusive Words
It is crucial for us to be aware, educated, and mindful of our use of words when discussing intelligence, whether in the context of AI or human capabilities. Misusing or oversimplifying psychological terms can lead to misunderstandings, perpetuate stereotypes, and hinder meaningful discussions. By striving to be accurate and informed, we can foster a more nuanced understanding of intelligence and its manifestations.
In the following weeks, we will explore other buzzwords from the realms of psychology or psychiatry that require careful consideration when used in the tech discourse.
If you have any questions or thoughts, please don't hesitate to reach out!
Until next time,
- Auxane Boch
References and Interesting Reads
Ackerman, P. L. (2023). Intelligence… moving beyond the lowest common denominator. American Psychologist, 78(3), 283.
APA Dictionary of Psychology. (n.d.). https://dictionary.apa.org/intelligence
Gardner, H. E. (2011). Frames of mind: The theory of multiple intelligences. Basic books.
Horn, J. L., & Cattell, R. B. (1966). Refinement and test of the theory of fluid and crystallized general intelligences. Journal of educational psychology, 57(5), 253.
Chicago
