What is Artificial Intelligence? Different definitions and applications?
This chapter describes the concept of Artificial Intelligence (AI) from various fields of study, providing the reader with a simple yet comprehensive understanding of the concept. Definitions are also presented to create a link between brain function and AI systems.
Artificial Intelligence (AI) is defined, from a computer science perspective, as a discipline capable of performing procedures similar to those of human intelligence (Estupiñán, Leyva, Peñafiel, and Ojeda, 2021). According to these authors, AI is conceived as “a set of technologies with characteristics or capabilities that were previously exclusive to the human brain”. According to these authors, AI is conceived as “a set of technologies with characteristics or capabilities that were previously exclusive to the human brain”. What’s more, this system incorporates two key approaches for its development: machine learning and deep learning.
Machine learning is a process in which systems recognize patterns, rapidly analyze large datasets, and make statistical predictions. Even machines designed with automatic thinking capabilities can now produce accurate models based on available data. Machine learning, on the other hand, focuses its algorithms on a specific variable, aiming to learn and predict future events (Coeckelbergh, 2020).
Deep learning has its origins in artificial neural networks, which are characterized by their high performance. In the field of cybersecurity, this technique is beneficial when dealing with diverse cases or large volumes of security data.
For example, deep learning can employ hybrid security models (understood as efficiency and reasoning ability (Lygeros, 2004)) to intelligently handle problems related to cybercrime, intrusion detection, malware analysis, security threat investigation, cyberattack prediction, and anomaly identification (Quirumbay, Castillo, and Coronel, 2022). In addition, Yang (2018) notes that deep learning provides significant advantages in building security models, thanks to its high accuracy in processing large datasets.
In a similar vein, Steels (2017) describes artificial intelligence as a scientific and engineering discipline focused on developing systems capable of mimicking certain functions, such as pattern recognition and language processing. He points out, however, that AI does not seek to reproduce human intelligence faithfully, but rather to solve specific problems that require a certain degree of intelligence. He suggests that, in the workplace, for example, its implementation should be carefully monitored and managed, given the impact it can have on both employees and employers.
From a psychological point of view, Torres (2014) and Icarte (2016) explain that artificial intelligence aims to replicate human cognitive functions, including problem-solving and decision-making. This approach, influenced by cognitive psychology and cognition theory, has guided the development of computer models designed to simulate complex human brain processes.
According to Piaget’s constructivist theory, knowledge is constructed through the individual’s interaction with cognitive and social factors, with experience playing a fundamental role in its development (Zambrano et al., 2016).
From a psychological point of view, Torres (2014) and Icarte (2016) explain that artificial intelligence aims to replicate human cognitive functions, including problem-solving and decision-making. This approach, influenced by cognitive psychology and cognition theory, has guided the development of computer models designed to simulate complex human brain processes.
According to Piaget’s constructivist theory, knowledge is constructed through the individual’s interaction with cognitive and social factors, with experience playing a fundamental role in its development (Zambrano et al., 2016). This concept also applies to the creation of intelligent systems, which, by being endowed with initial knowledge such as the data with which the individual was born and by interacting with the user, can learn and acquire new knowledge over time, i.e., to learn through experience.
Artificial neural networks, computer models capable of learning and solving problems, are a good example. These networks, when interacting with new data, adjust their connections and improve their performance, thereby acquiring knowledge through experience (Magiquo, 2019).
Russell and Norvig (2004) provide an interesting definition that helps us understand Figure 1. They consider AI to be “the study of agents that receive perceptions from the environment and perform actions”. In this context, each agent plays a role sequentially, resulting in a variety of actions. These roles are represented in “production systems, reactive agents, real-time planners, neural networks, and decision-theoretic systems”.
AI is now described as an algorithm designed to simulate a form of intelligence, which, in turn, is incorporated into digital platforms, social networks, and search engines, positioning the technology as an agent that plays a crucial role in the way we communicate and interact.
The hypothesis receives empirical support, given that users not only consume data but also constantly produce it, thereby contributing to the massive generation of information in all fields (Coeckelbergh, 2020).
References
- Aanbevolen citeerwijze: Luc Steels (ed.), Artificiële intelligentie. Naar een vierde industriële revolutie? KVAB Standpunt 53, 2017.
- Coeckelbergh, Mark. (2020). AI ethics. Hoe omgaan met artificiële intelligentie?
- Estupiñán Ricardo, J., Leyva Vázquez, M. Y., Peñafiel Palacios, A. J., & El Assafiri Ojeda, Y. (2021). Inteligencia artificial y propiedad intelectual. Revista Universidad y Sociedad, 13(S3), 362-368.
- Icarte Ahumada, G. A. (2016). Aplicaciones de inteligencia artificial en procesos de cadenas de suministros: una revisión sistemática. Ingeniare. Revista chilena de ingeniería, 24(4), 663-679.
- John Lygeros, Lecture Notes on Hybrid Systems, Department of Electrical and Computer Engineering – University of http://robotics.eecs.berkeley.edu/~sastry/ee291e/lygeros.pdf
- Quirumbay Yagual, D. I., Castillo Yagual, C. A., & Coronel Suárez, I. A. (2022). Una revisión del aprendizaje profundo aplicado a la ciberseguridad. Revista Científica y Tecnológica UPSE (RCTU), 9(1), 57-65.
- Torres, C. E. T. (2014). Inteligencia colectiva: enfoque para el análisis de redes. Estudios Gerenciales, 30 (132), 259-266.
- Russell, S.J. y P. Norvig (2004). Inteligencia artificial. Un enfoque moderno. Pearson Educación, S.A.
- XIN, Yang; KONG, Lingshuang; LIU, Zhi; CHEN, Yuling; LI, Yanmiao; ZHU, Hongliang; GAO, Mingcheng; HOU, Haixia y WANG, Chunhua. Machine Learning and Deep Learning Methods for Cybersecurity. IEEE Access [En línea]. 2018, vol. 6, pág. 35365-35381. ISSN 21693536. Disponible en: 10.1109/ACCESS.2018.2836950
- Zambrano, P. J. S., Cedeño, G. D. R. B., & Rivadeneira, M. R. L. (2016). La teoría constructivista de Jean Piaget y su significación para la pedagogía contemporánea. Dominio de las Ciencias, 2(3), 127-137.
Image references
- Magiquo. (2019, noviembre 1). Redes neuronales o el arte de imitar el cerebro humano. Magiquo. https://magiquo.com/redes-neuronales-o-el-arte-de-imitar-el-cerebro-humano/
0 Comments