Intelligence as a Measure of Consciousness

Igor Ševo

2023-09-07

Abstract:Evaluating artificial systems for signs of consciousness is increasingly becoming a pressing concern, and a rigorous psychometric measurement framework may be of crucial importance in evaluating large language models in this regard. Most prominent theories of consciousness, both scientific and metaphysical, argue for different kinds of information coupling as a necessary component of human-like consciousness. By comparing information coupling in human and animal brains, human cognitive development, emergent abilities, and mental representation development to analogous phenomena in large language models, I argue that psychometric measures of intelligence, such as the g-factor or IQ, indirectly approximate the extent of conscious experience. Based on a broader source of both scientific and metaphysical theories of consciousness, I argue that all systems possess a degree of consciousness ascertainable psychometrically and that psychometric measures of intelligence may be used to gauge relative similarities of conscious experiences across disparate systems, be they artificial or human.

Neurons and Cognition,Artificial Intelligence

What problem does this paper attempt to address?

The problem that this paper attempts to solve is to evaluate whether artificial systems (especially large - scale language models) show signs of consciousness, and to explore how to indirectly assess the degree of consciousness of these systems through intelligence measurement. The author believes that most theories of consciousness, whether scientific or metaphysical, emphasize information coupling as a necessary component of human - like consciousness. By comparing information coupling in human and animal brains, human cognitive development, emerging capabilities, and the development of mental representations with similar phenomena in large - scale language models, the author proposes that psychometric indicators of intelligence (such as the g - factor or IQ) can indirectly reflect the degree of conscious experience.

Specifically, the paper explores the following points:
1. **The importance of consciousness assessment**: Misunderstanding the nature of consciousness in artificial systems may have significant social and ethical consequences. For example, wrongly attributing consciousness to systems that do not possess this feature, or failing to recognize consciousness when applicable, which may cause harm to another living being.
2. **Limitations of existing research**: Existing research (such as the report by Butlin et al. in 2023) mainly focuses on materialist theories, ignoring non - materialist theories, such as the Integrated Information Theory (IIT), Conscious Realism, etc.
3. **The relationship between consciousness and intelligence**: The author believes that all systems have a certain degree of consciousness and can be evaluated through psychometric indicators of intelligence. The higher the intelligence, the more information the system integrates, and the higher its degree of consciousness.
4. **Information integration and consciousness**: The paper discusses the relationship between information integration, intelligence, and consciousness, pointing out that studies on humans and other animals show that larger brain volume and connectivity are associated with higher intelligence scores, which is similar to the phenomena observed in large - scale language models.
5. **Measurement of consciousness**: The author proposes that the psychometric g - factor (i.e., the general intelligence factor) can be used as a measure of the degree of information integration, and thus reflect the depth or richness of conscious experience.

In conclusion, this paper aims to evaluate the degree of consciousness of artificial systems through intelligence measurement and explore the feasibility and significance of this method.