Artificial intelligence paradigms

Artificial intelligence paradigms

1. When it comes to Artificial Intelligence everything is context. When they ask me how the technology works, I tell them that it is very simple: interpretation of stimuli, processing, qualification, most likely response and learning. Its simplicity is matched by its complexity of setup, training, and adoption. The world of AI is usually not very accommodating and could not be different. Recently there was talk of a concept: "artificial stupidity." According to experts from the Carlos III University of Madrid, the algorithms had been designed to solve specific problems, but not to understand what is happening in their environment, they only see signals and always apply the same solution. The machines are totally autistic and giving them an understanding of their environment is very complicated, specialists argue. An example of this are some artificial vision systems that could mistake a toy turtle for a rifle and black people for apes. These devices have been trained to see patterns and

   • COMPUTATIONAL MODELS Science has historically been based on models to represent the universe and especially the real world. We can say that artificial intelligence has always sought a model that represents human intelligence, but the reality has been different, since there is no single model capable of representing intelligence, there is even no definition of artificial intelligence that meets expectations of all researchers. In computing, the first theories and models were inspired by the representation of the mind. Since artificial intelligence started from the principle of intelligence, it could be represented by means of algorithms to be executed on a computer. We can generally classify the models into six: Symbolic model – Logical: production rules or semantic: concepts Connectionist or emerging model – Neural networks. Data-driven (operational) model – Deep learning. Collective model – Collective intelligence Evolutionary model – Genetic algorithm. Corporeal model – Reactive age

2. What is Artificial Intelligence and what is it used for? Artificial Intelligence is defined as that technology capable of performing operations comparable to those of the human mind. In business terms, its main applications are based on compression and problem solving; that is, this intelligence has the ability to analyze the situation and make decisions. Its arrival is a way to increase productivity and improve processes in work environments, generating more competitive production paradigms, as well as being one of the main drivers of transformation. Artificial Intelligence allows knowing relevant aspects about the company's operation, as well as the market in which it operates in order to identify possible risks and find new growth opportunities through the analysis of the data obtained. At Aura we understand the importance of optimizing business processes and results. For this reason, we offer our clients solutions that help boost all areas of their business. We ensure that our cl

   • In its beginnings, there was talk of a symbolic paradigm that sought to represent thought as symbols that could be converted into programs and executed on the computer where the programs represented the mental states and the computer the functioning of the brain. Here we use the term cognitive paradigm for both the symbolic and the connectionist models since both start from the reproduction of the functions of the mind or brain in a computer. When we speak of mind or brain, we are speaking of an individual, that is, of the states of a mind or of the functions of a brain, which can be represented in the form of algorithms and be executed in a machine. For some, humans and computers are two species of the same genus: information processing systems. Accepting the human mind as a form of information processing results in two inverse phenomena: The generalization of information processing systems. The simplification of the human mind (reductionism). The problem lies in the generalization

3. From the crisis of the symbolist paradigm a new paradigm arises within artificial intelligence: the connectionist paradigm that assumes that intelligence is the product of the association, combination or connection of simpler independent entities (agents, genes or neurons), which they may have some or no intelligence and as a result of their interaction, the system would present an intelligent behavior (Towards a new approach to artificial intelligence). AI did not stay in the connectionist or emergent paradigm, which has been followed by the enactive or corporeal paradigm and the paradigm based on data or superintelligent machines. For now, we can mention the following paradigms Symbolic paradigm (logical/semantic) Connectionist or emerging paradigm corporeal or enactive paradigm Data-driven paradigm or super-intelligent machines In turn, the paradigms have given rise to new models and new research techniques (machines surpassing humans) among the most important techniques we h

   • SYMBOLIC MODEL It was the first of the information representation models used by AI and is based on symbol processing systems. Symbol processing system: System capable of manipulating symbols, generating new symbols, creating and modifying relations between symbols, storing symbols, etc. The symbols are physical and have a physical-electronic substrate. Where the substrates can be: Physical-electronic (computer) Physical-biological (human beings) The adjective "physical" denotes that it is the material realization of a formal system. Therefore, a physical symbol system is a formal system executed by a real machine, generally in a computer where the brain would be another type of machine. where intelligence is produced by executing a system of symbols These are the elements that gave rise to the concept of a strong AI, all that is missing is a hypothesis that would legitimize it. Theoretical Assumptions Symbolic model hypothesis: the nature of the substrate (electronic or biological