Traditionally, artificial intelligence (AI) studies assumed that the human brain manipulates symbols, and the models that these studies suggested for intelligency were based on algorithms manipulating symbols. Knowledge was represented explicitly. Negative aspects of these models were the fact that they could handle just coerent knowledge, while human knowledge is characterised by not being coerent. Models implemented to solve problems had to deal with combinatory explosion. Nowadays, AI focuses on cognitive systems using their intelligence to adapt to its environment. These systems still use symbolic representations, but these representations are not explicited, they emerge spontaneously from subsymbolic computation.

Subsymbolism

Subsymbolism means that concepts, instead of being represented with a symbol (number, letter, word, phrase, memory adress, etc.), are represented with quantitive properties. The single quantitive property is just a sub-part of the representation of the concept, a subsymbol. Traditional AI assumes that our reasoning consists in handling symbols representing concepts, and identifies these symbols with our natural language. Subsymbolism is a valid alternative to this assumption. Concerning neural networks, the subsymbolism paradigm stresses the importance of distributing representations on multiple units. This means for example that, if a concept has a certain property, this property is not represented by the activation of one unit, but by the activation of multiple units.

Biologic inspiration

Recent developments in AI introduced a new way to approach intelligence. New methods are inspired on biologic structures and processes:

• connectionist systems (neural networks) inspire their architecture on the brain architecture using neuronal units and dendrite-like connections between them
• genetic algorithms are inspired on the evolutionary laws of Darwin and Lamarck.
• Multiple agent systems focuse on social interaction

Emergentist and dualistic vision

The dualistic vision states that mental activity takes place on a different level than the level of the physical structures of the brain and its environment, and that these two levels are scarsely interdependent. The emergent vision, instead, states that mental activity emerges from the physical structures, and is strictly dependent from them. Mental activity is a result on macroscopic level of the activity on the microscopic level of physical brain structures. The symbolic level emerges from the subsymblic level. One important distinction between traditional AI and today's AI is that, while in traditional AI systems the symbolic level is the only level present, in connectionist systems the symbolic level is absent when the system is in its initial state (state 0 in dynamics theory), and emerges from the subsymbolic level when the system creates its representations. Emergentist and dualistic vision are not new phenomena, Aristoteles can be considered an exponent of the emergentist vision, while Plato shared the dualistic vision.