Abstract
Success in all sorts of situations is the most classical interpretation of general intelligence. Under limited resources, however, the capability of an agent must necessarily be limited too, and generality needs to be understood as comprehensive performance up to a level of difficulty. The degree of generality then refers to the way an agent’s capability is distributed as a function of task difficulty. This dissects the notion of general intelligence into two non-populational measures, generality and capability, which we apply to individuals and groups of humans, other animals and AI systems, on several cognitive and perceptual tests. Our results indicate that generality and capability can decouple at the individual level: very specialised agents can show high capability and vice versa. The metrics also decouple at the population level, and we rarely see diminishing returns in generality for those groups of high capability. We relate the individual measure of generality to traditional notions of general intelligence and cognitive efficiency in humans, collectives, non-human animals and machines. The choice of the difficulty function now plays a prominent role in this new conception of generality, which brings a quantitative tool for shedding light on long-standing questions about the evolution of general intelligence and the evaluation of progress in Artificial General Intelligence.
Highlights
Under the interpretation of generality we have introduced in this paper, a general system would have a strong bias in favour of easy problems
When we look at an agent characteristic curve, the notions of capability and generality appear as the two most descriptive indicators to summarise the curve
Generality can increase or decrease individually or collectively, through evolution, development or design
Summary
The first scientific notion of general intelligence for humans was introduced by Charles Spearman[6]. The test correlations were positive, something that is known as a ‘positive manifold’ He derived the notion of the ‘g factor’, which is a latent factor that explained the variability in the matrix. This concept of the g factor is populational: the g factor does not provide a measure of the generality of an agent, but a feature of the population sample. From this initial view of a dominant g factor explaining most of human cognitive performance, theories of intelligence have taken less extreme views, such as Carroll’s three-stratum t heory[22] or Sternberg’s triarchic theory of intelligence[23]. While instance difficulty plays an important role in IRT, the distinction between ability and generality and the connection between generality and task difficulty have not been made explicit to date
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
