Navigating the Complexities of Intelligence: Conceptions, Implications, and Challenges

Something as complex and amorphous as intelligence cannot be easily defined. Many conceptions of intelligence have been put forward across various intellectual disciplines throughout recorded history, ranging from ancient philosophy to contemporary science. Gaining prominence in the writings of Plato, who believed intelligence to be the faculty that distinguished social classes (Gençten, 2018), the Ancient Greek term for intelligence is phronesis, denoting practical action, good judgement, and excellence of character (Blackburn, 2008). The predominant conception of intelligence as a measurable construct, commonly known as psychometric intelligence, started gaining prominence in the late 19th century. Most theories developed in the last 100 years agree that intelligence reflects a general ability to accomplish goals in a wide range of environments (Legg & Hutter, 2007). Differences in intelligence come with serious practical consequences: high intelligence is positively correlated with exceptional performance in key life domains such as educational, occupational, and societal status, which are in turn associated with positive long-term health outcomes, while low intelligence is negatively correlated with those same factors (Plomin & Deary, 2015), suggesting that differences in intelligence bring serious, life-altering consequences. Although intelligence is not the only determining factor in broad life outcomes, this essay will argue that differences in intelligence shape social mobility and opportunities, impacting the lives of individuals, their families and the communities in which they live.

Psychologists working in the field no longer debate the existence and structure of intelligence. While many theories of intelligence have been proposed, central to the literature is a general cognitive factor common to many mental abilities known as g, or general intelligence (Spearman, 1904). General intelligence can be defined as a broad cognitive capability that involves the ability to comprehend complex ideas, solve problems, and learn from experience (Gottfredson, 1997). Every study that has applied a mixed battery of cognitive tests to a large enough sample size in the last century has found evidence of this same shared cognitive variance, or g, across different testing mediums (pencil & paper, one-to-one), content (numerical, spatial, verbal), and function (reasoning, memory, vocabulary), with correlations as high as .90 (Deary, 2001; Jensen, 1998 as cited in Deary, 2001). However, cultural biases in the design of these tests may have led to overlooking aspects of intelligence more relevant in other cultures. The score on cognitive ability tests is generally referred to as an intelligence quotient (IQ), with g acting as an index for IQ, and these scores are usually converted to a scale in which the mean (M) is 100 and the standard deviation (SD) is 15 (Neisser et al., 1996). Although there are many types of IQ tests, they almost always correlate positively, meaning those who score well on one mental task tend to score well on others (Plomin & Deary, 2015), indicating the existence of an underlying capacity. Having discussed the structure of intelligence in some detail, we are now in a position to consider some of its real-world implications.

Intelligence is one of the strongest predictors of occupational success, particularly in highly complex roles. A review of 85 years of research in personnel selection found IQ to be a powerful predictor of job success with an average correlation of .50, suggesting a causal link between high intelligence test scores, rapid assimilation and application of job-related knowledge, and, subsequently, high performance (Schmidt & Hunter, 1998). This indicates that a core aspect of g involves adaptability: comprehending complex ideas and quickly learning from experience. Multiple lines of evidence support this view, finding that people with higher intelligence test scores typically scan, retrieve, and respond to stimuli more quickly than those with lower intelligence test scores (Neisser et al., 1996). However, the fact that there is not a complete correlation (1.0) between intelligence and job success suggests that occupational success does not solely rely upon intelligence and may be mediated by other factors not easily measured by intelligence tests like personality traits, social intelligence, and creativity.

General intelligence grows in predictive validity as task complexity increases, meaning that g correlates with better performance when higher levels of cognitive ability are required. The United States (US) military measures cognitive ability as part of its induction process. Because periods of history necessitated training inductees of all aptitudes, such as during World War II, the US military has devoted much research to general intelligence training, thereby providing evidence for its causal importance. By attempting to train the cognitive skills associated with general intelligence, including reading comprehension, to recruits with lower intelligence scores, the US military found that those with lower intelligence scores could at most increase performance to minimally acceptable levels in positions that do not require high-order thinking abilities concluding that, even with extended training windows, their intelligence training efforts were largely unsuccessful (Gottfredson, 1997). In this same vein, work by Vineberg and Taylor (1972 as cited in Gottfredson, 1997) demonstrated that it took lower aptitude recruits (as defined by IQs between the 10th to 30th percentile) 12 to 24 months to perform at a level similar to which higher aptitude recruits (30th percentile and above) exhibit after no more than 3 months of experience, showing that individual differences do not decrease with practice, particularly in tasks that are not easily systematised. This indicates that tasks requiring the comprehension and application of inconsistent information require higher levels of general intelligence. However, because these results are 50 years old, it is unclear whether these would apply in the modern technologically-driven workplace. As a consequence of these findings, the military is prohibited by law from inducting individuals below the 10th percentile (IQ 80), partly because the resources needed to train these individuals to perform at minimally acceptable standards outweighed the benefits of enlistment (Gottfredson, 1997). Given the harsh Darwinian nature of combat, where larger brigades bring a competitive advantage, the rejection of those with those below IQ 80 is a powerful example of the causal significance of general intelligence. Although it is not the sole predictor of professional achievement, the evidence outlined above suggests that intelligence plays a crucial role in occupational success.

The influence of intelligence extends beyond occupational outcomes to other areas of social life. The growing complexity of contemporary society disproportionately benefits those with higher intelligence, creating social inequalities based on IQ. While the old industrial economy rewarded mass production and standardisation (Gottfredson, 1997), modern post-industrial economy rewards complex services such as enterprise software products built by multidisciplinary teams of highly educated individuals making fast-paced decisions in unison. Society is becoming increasingly automated, displacing blue-collar jobs (Helbing, 2015), while simultaneously putting an increasing emphasis on highly abstract professions such as software development, robotics engineering, and scientific research that require complex higher-order thinking. For example, many modern car manufacturing factories are almost completely automated, with operations overseen by highly specialised engineers (Helbing, 2015). Many of the duties that correlate highly with overall job complexity now spill over into day-to-day life: advising, planning, negotiating, and supervising; self-service banking and store checkouts force individuals to wrestle with their own capabilities (Gottfredson, 1997). As a result, individuals with below-average intelligence are having trouble adapting to modern life (Granat & Granat, 1978), and if left unaddressed, may increase social inequality due to the inability to cope with increased complexity. Let us now consider another significant aspect of intelligence in the context of contemporary life: artificial intelligence (AI).

Recent advancements in AI challenge dominant conceptions of intelligence and further complicate intelligence disparities. AI is an umbrella term that refers to non-biological capacities to autonomously achieve complex goals (Korteling, 2021). In a recent review of 1,016 occupations within the US labour market, Eloundou and colleagues (2023) predicted that the next generation of AI—pioneered by market leader OpenAI—will disrupt around 80% of the US workforce, and 19% of these roles having 50% of concomitant tasks automated. As manufacturing automation drove the loss of blue-collar jobs, societal integration of AI is beginning to displace white-collar and highly skilled professional jobs: AI-powered tools outperform physicians in detecting diseases such as cancer and remove the need for large teams of lawyers to synthesise legal documents (Haenlein & Kaplan, 2019).

A concern in the discourse surrounding AI is the superiority of psychometric intelligence, potentially overlooking other forms of intelligence, such as social and emotional intelligence. This discrepancy has been addressed by Gardner’s theory of multiple intelligences (1983), which posits that intelligence goes beyond traditional psychometric measures, stressing the importance of linguistic, musical, and intrapersonal abilities, challenging the prevailing IQ-based model. Similarly, Moravec’s paradox notes that tasks that are simple for the ancient biology of human brains (e.g. face perception) are difficult for AI’s digital technology, implying that higher-order computation requires little computational resources compared to low-level perceptual-motor skills (Korteling, 2021). While human cognitive abilities underpinning what is typically associated with intelligence are less than 100,000 years old, our biological perceptual-motor skills have been developing for millions of years, facilitating the refinement of what are actually highly complex processes such as facial perception and socio-emotional attunement (Korteling, 2021), suggesting that humans possess underlying capabilities that might not be captured by g-focused measurements. Similarly, Huang et al. (2019) suggest that computationally-heavy tasks typically associated with intelligence will be taken over by AI, while humans will increasingly specialise in ‘feeling tasks’, involving our older, precognitive capacities, which require the capability to recognise, understand, and respond to often ambiguous emotions. However, this analysis could be too simplistic, as it fails to consider the potential for AI to develop social and emotional forms of intelligence. By exploring capacities outside of the traditional psychometric perspective, our understanding of intelligence can be enriched and broadened, facilitating a more complete account of the interplay between various cognitive and emotional capacities.

In conclusion, intelligence is a multifaceted construct that plays a pivotal role in shaping individual lives and societal structures. The implications of intelligence differences extend beyond academic and occupational success, reaching into the fabric of society, shaping social mobility, opportunities and the overall well-being of individuals and their communities. As society grows in complexity, partially driven by rapid technological innovation, the effects of intelligence disparities are amplified, consequently widening social inequalities. As AI evolves and continues to displace jobs previously filled by human counterparts, it is crucial to recognise and nurture the diverse forms of intelligence that, for now, are distinctly human, such as emotional and social intelligence. By adopting a more integrative approach that embraces and thrives on change, we can work towards a more equitable world where the full spectrum of intelligence—both biological and non-biological—is acknowledged and leveraged for the greater good of all.

References

Blackburn, S. (2008). The Oxford Dictionary of Philosophy (2nd ed.). Oxford University Press. https://doi.org/10.1093/acref/9780199541430.001.0001

Deary, I. J. (2001). Intelligence: A Very Short Introduction (First Edition). Oxford University Press.

Eloundou, T., Manning, S., Mishkin, P., & Rock, D. (2023). GPTs are GPTs: An Early Look at the Labor Market Impact Potential of Large Language Models (arXiv:2303.10130). arXiv. https://doi.org/10.48550/arXiv.2303.10130

Gençten, N. B. (2018). The Concept of Intelligence from Ancient Greek Philosophy to the Theory of Multiple Intelligences. https://www.academia.edu/36815479/The_Concept_of_Intelligence_from_Ancient_ Greek_Philosophy_to_the_Theory_of_Multiple_Intelligences

Gottfredson, L. S. (1997). Why g matters: The complexity of everyday life. Intelligence, 24(1), 79–132. https://doi.org/10.1016/S0160-2896(97)90014-3

Granat, K., & Granat, S. (1978). Adjustment of intellectually below-average men not identified as mentally retarded. Scandinavian Journal of Psychology, 19(1), 41–51. https://doi.org/10.1111/j.1467-9450.1978.tb00302.x

Haenlein, M., & Kaplan, A. (2019). A Brief History of Artificial Intelligence: On the Past, Present, and Future of Artificial Intelligence. California Management Review, 61(4), 5–14. https://doi.org/10.1177/0008125619864925

Helbing, D. (2015). The Automation of Society is Next: How to Survive the Digital Revolution. https://doi.org/10.2139/ssrn.2694312

Gardner, H. (1983). Frames of Mind: The Theory of Multiple Intelligences. Basic Books.

Huang, M.-H., Rust, R., & Maksimovic, V. (2019). The Feeling Economy: Managing in the Next Generation of Artificial Intelligence (AI). California Management Review, 61(4), 43–65. https://doi.org/10.1177/0008125619863436

Korteling, J. E. (Hans)., van de Boer-Visschedijk, G. C., Blankendaal, R. A. M., Boonekamp, R. C., & Eikelboom, A. R. (2021). Human- versus Artificial Intelligence. Frontiers in Artificial Intelligence, 4. https://www.frontiersin.org/articles/10.3389/frai.2021.622364

Legg, S., & Hutter, M. (2007). Universal Intelligence: A Definition of Machine Intelligence. https://doi.org/10.48550/ARXIV.0712.3329

Neisser, U., Boodoo, G., Bouchard Jr., T. J., Boykin, A. W., Brody, N., Ceci, S. J., Halpern, D. F., Loehlin, J. C., Perloff, R., Sternberg, R. J., & Urbina, S. (1996). Intelligence: Knowns and unknowns. American Psychologist, 51, 77–101. https://doi.org/10.1037/0003-066X.51.2.77

Plomin, R., & Deary, I. J. (2015). Genetics and intelligence differences: Five special findings. Molecular Psychiatry, 20(1), Article 1. https://doi.org/10.1038/mp.2014.105

Schmidt, F. L., & Hunter, J. E. (1998). The validity and utility of selection methods in personnel psychology: Practical and theoretical implications of 85 years of research findings. Psychological Bulletin, 124, 262–274. https://doi.org/10.1037/0033-2909.124.2.262

Spearman, C. (1904). ‘General intelligence,’ objectively determined and measured. The American Journal of Psychology, 15, 201–293. https://doi.org/10.2307/1412107

Sternberg, R., & Sternberg, K. (2016). Cognitive Psychology (7th edition). Wadsworth Publishing.