Cognitive architectures can be symbolic, connectionist or hybrid.[7] Some cognitive architectures or models are based on a set of generic rules, such as, for example, the Information Processing Language (for example, Soar based on the unified theory of cognition, or similarly ACT-R). Many of these architectures are based on the analogy that the mind is like a computer. In contrast, subsymbolic processing does not specify such a priori rules and relies on emergent properties of processing units (e.g., nodes). Hybrid architectures combine both types of processing (such as CLARION "CLARION (cognitive architecture)"). A further distinction is whether the architecture is centralized with a neural correlate of a processor at its core or decentralized (distributed). The decentralized approach has become popular under the name parallel distributed processing in the mid-1980s and connectionism, a prime example being neural networks. Another design issue is also the choice between a holistic and atomistic structure or a modular (more concrete) structure.

In contrast to relatively static symbolic architectures such as Soar, the OAGI Ontogenetic Architecture of General Intelligence proposes a phased ontogenetic development starting from an initially undifferentiated substrate that specializes into functional modules through processes of self-organization, computational morphogenesis, and embodied learning. OAGI places special emphasis on sensorimotor embodiment, critical windows of plasticity, and explicit mechanisms of ethical governance during development; It is thus presented as an alternative approach that prioritizes gradual growth and adaptive plasticity over the a priori definition of fixed symbolic structures.

In traditional AI, intelligence is often programmed from the top: the programmer is the creator, creating something and imbuing it with his intelligence, although many traditional AI systems were also designed to learn (for example, improving their proficiency in games or problem solving). Biologically inspired computing, on the other hand, sometimes takes a more decentralized and bottom-up approach; Bioinspired techniques often involve the method of specifying a set of simple generic rules or a set of simple nodes, from whose interaction general behavior emerges. It is expected to accumulate complexity until the final result is something markedly complex (see complex systems). However, it can also be argued that systems designed top-down based on observations of what humans and other animals can do rather than observations of brain mechanisms are also biologically inspired, although in a different way.

Some known cognitive architectures, in alphabetical order:.

Cognitive architectures can be symbolic, connectionist or hybrid.[7] Some cognitive architectures or models are based on a set of generic rules, such as, for example, the Information Processing Language (for example, Soar based on the unified theory of cognition, or similarly ACT-R). Many of these architectures are based on the analogy that the mind is like a computer. In contrast, subsymbolic processing does not specify such a priori rules and relies on emergent properties of processing units (e.g., nodes). Hybrid architectures combine both types of processing (such as CLARION "CLARION (cognitive architecture)"). A further distinction is whether the architecture is centralized with a neural correlate of a processor at its core or decentralized (distributed). The decentralized approach has become popular under the name parallel distributed processing in the mid-1980s and connectionism, a prime example being neural networks. Another design issue is also the choice between a holistic and atomistic structure or a modular (more concrete) structure.

In contrast to relatively static symbolic architectures such as Soar, the OAGI Ontogenetic Architecture of General Intelligence proposes a phased ontogenetic development starting from an initially undifferentiated substrate that specializes into functional modules through processes of self-organization, computational morphogenesis, and embodied learning. OAGI places special emphasis on sensorimotor embodiment, critical windows of plasticity, and explicit mechanisms of ethical governance during development; It is thus presented as an alternative approach that prioritizes gradual growth and adaptive plasticity over the a priori definition of fixed symbolic structures.

In traditional AI, intelligence is often programmed from the top: the programmer is the creator, creating something and imbuing it with his intelligence, although many traditional AI systems were also designed to learn (for example, improving their proficiency in games or problem solving). Biologically inspired computing, on the other hand, sometimes takes a more decentralized and bottom-up approach; Bioinspired techniques often involve the method of specifying a set of simple generic rules or a set of simple nodes, from whose interaction general behavior emerges. It is expected to accumulate complexity until the final result is something markedly complex (see complex systems). However, it can also be argued that systems designed top-down based on observations of what humans and other animals can do rather than observations of brain mechanisms are also biologically inspired, although in a different way.

Some known cognitive architectures, in alphabetical order:.