Intelligence in nature is very diverse; it can be discussed in terms of the intelligence of chimps, dolphins, and many other animals and even the collective intelligence of very simple life forms such as ants. The concept of intelligence is being adopted in many domains of science.

From practical and theoretical perspectives, an essential current research direction that is under examination by a very large number of scientists worldwide is the development of artificial systems, also called agent-based systems, which can be either individual agents or cooperative multiagent systems. These systems are embedded in the environment, possess a certain degree of autonomy, and are capable of perceiving the environment and executing actions in it. Intelligence in agent-based systems can emerge through advanced problem-solving abilities. To develop these intelligent systems, techniques such as supervised learning and unsupervised learning play pivotal roles. Supervised learning involves training systems on labeled datasets, enabling them to map inputs to outputs and improve their predictions or classifications over time. These methods are extensively used in applications such as image recognition, natural language processing, and medical diagnosis. On the other hand, unsupervised learning enables systems to identify patterns and structures in unlabeled data, facilitating clustering, anomaly detection, and exploratory data analysis. A more recent paradigm, retrieval-augmented generation (RAG) and cache-augmented generation (CAG), combines retrieval- and cache-based methods with generative models to enhance intelligent systems, producing highly contextualized and accurate responses or actions. These approaches are instrumental in developing advanced systems for domains such as conversational AI, knowledge management, and decision support. Intelligent agent-based systems have many real-world applications, including the health sciences and industry. The number and diversity of intelligent systems are increasing rapidly. In this context, the great challenge consists of creating increasingly intelligent systems. Another critical challenge, approached by only a few researchers worldwide, is the development of universal metrics, including black-box-based intelligence metrics, to measure the intelligence of these systems. Such advances could allow for the comparison of systems based on their intelligence. In cooperative multiagent systems, intelligence can be assessed at the system level; this is known as collective intelligence. Even in very simple cooperative multiagent systems, agents may interact nonlinearly at various decision points, resulting in emergent complexity and intelligence at the system level.