This article aims to critically analyze the epistemological foundations of artificial intelligence (AI). By examining works that explain how this technology works, it is understood that its epistemological basis is made up of the inductive method and statistics based on a mathematization of reality. It is these elements that allow machines to learn by recognizing patterns and to make predictions and provide answers. However, these foundations have limitations and problems that have been discussed by philosophers throughout history. In this (...) article we will present how induction and mathematization function as the epistemological basis of artificial intelligence and how some of the limitations of this technology can be explained through the weaknesses of the methods that support it. (shrink)

O uso da inteligência artificial no Poder Judiciário ainda encontra fortes limitações à luz da compreensão representacional-computacional da mente. A exemplo do projeto pioneiro do juiz robô na Estônia, a dificuldade reside na incapacidade de apreensão da narrativa subjetiva da máquina. Logo, através da experiência do ser-aí-no- mundo de Heidegger, define-se um contraste entre cognição humana e funcionamento computacional. Partindo de uma pesquisa qualitativa e de cunho bibliográfico, busca-se demonstrar a complexidade de conceber dispositivos puramente físicos com inteligências aptas (...) a substituírem juízes por robôs. Conclui-se que ainda não há tecnologia capaz de substituir a ciência cognitiva heideggeriana do Dasein. (shrink)

El presente artículo propone un modelo de análisis para orientar el discernimiento y posicionamiento en torno al uso de la IA en educación superior. Analizaremos los cuatro posibles usos de la IA que más incidencia tendrán en esta si se considera el actual desarrollo de esta tecnología. Ellos son: la co-construcción de conocimiento, el uso de IA para tutoría o mentoreo, el auge de las plataformas adaptativas y la capacidad predictiva. Propondremos un ejercicio de análisis para cada uso apelando a (...) algunas de las dimensiones del modelo propuesto para orientar sobre su utilización. El resultado es fecundo, en tanto advierte sobre el riesgo de parcializar el análisis perdiendo de vista las implicancias y los compromisos más profundos que exige el discernimiento sobre el uso de la IA. (shrink)

Las investigaciones que Carlos Alchourrón desarrolló en la Filosofía del Derecho se vincularon, desde inicios de la década de 1980, con problemáticas críticas de la Inteligencia Artificial. Su contribución a la construcción de una lógica de las normas se conectó rápidamente con la deducción automática y los Sistemas Expertos Jurídicos. Su preocupación por la cuestión de los conflictos de obligaciones que pueden plantearse en un sistema normativo cuando el juez se enfrenta a la necesidad de emitir un veredicto (...) considerando ciertos hechos, se vinculó, en sus trabajos, con la noción más general de "condicionales derrotables" y con la formalización del Razonamiento No Monótono. Pero su contribución de mayor impacto fue aquella derivada de la búsqueda de una solución al problema de los resultados múltiples que surgen de la derogación de una norma en un sistema jurídico, problema que puede ser visto, en un contexto epistémico, como el de la contracción de un conjunto de creencias. Su enfoque constructivo, que desarrolló con David Makinson, convergió con el axiomático elaborado por Peter Gärdenfors, dando lugar al surgimiento de la teoría de los Cambios Racionales de Creencias, más conocida como AGM, teoría que fue adoptada en IA como un modelo standard para la especificación de las actualizaciones de Bases de Conocimiento, operaciones que, habitualmente, involucran la introducción de una nueva creencia que puede ser inconsistente con las anteriores. ¿Por qué AGM adquirió rápidamente carta de ciudadanía en IA, al punto de que su influencia, medida según la cantidad de menciones bibliográficas del artículo fundacional de 1985, se ha duplicado en la última década? Mostraremos en este trabajo que la teoría apareció en un momento de crisis de la IA y que el alto nivel de abstracción del modelo de AGM era lo que muchos investigadores de la disciplina estaban buscando. El presente artículo, luego de describir la génesis de AGM, realiza un análisis histórico de las circunstancias en las cuales la teoría se introdujo en la IA y una evaluación cuantitativa y cualitativa de su impacto en dicha disciplina. The main lines of research that Carlos Alchourrón developed in the Philosophy of Law, were connected, from the beginnings of the 80' s, to some critical problems in the Artificial Intelligence. His contributions to the construction of a logic of norms were rapidly linked to automatic deduction and the Legal Expert Systems. His concerns for the conflict of obligations that could arise in a normative system when a judge has to issue a ruling taking some particular facts into account were associated in his work with the more general notion of "defeasible conditionals" and with the formalization of Non Monotonic Reasoning. But his most relevant contribution was the one derived from the search of a solution to the problem of multiple outcomes resulting from the derogation of a norm in a legal corpus, which may be considered -within an epistemic framework- analogous to the contraction of a set of beliefs. The constructive approach he worked out with David Makinson converged with Peter Gärdenfors' axiomatic, giving rise to the formulation of the theory of Rational Changes of Belief, which was adopted in AI as a model for the specification of the updates of Knowledge Bases as it usually involves a new belief that may be inconsistent with the old ones. (shrink)

The aim of this paper is to study the temporal de-synchronization that occurs, within the European Union (EU), between technological innovations in Artificial Intelligence and its legal regulations. To do that, I base on a thesis formulated by Hartmut Rosa, a German sociologist close to the Critical Theory of society, according to which the temporal structures of politics are no longer compatible with the pace of change in some others social spheres. This produces a new form of social pathology, (...) which Rosa calls "risks of de-synchronisation", and whose most worrying effect must be sought in the displacement of decision-making processes from the political sphere to faster areas of society, fundamentally the market. To analyze this problem, I first reconstruct the thesis of desynchronization as a social pathology derived from acceleration. Then, I map the AI legislation process at EU level. Finally, I defend the thesis that this legislative process, which is certainly very admirable, has consistently come too late. El objetivo de este trabajo es estudiar el desajuste temporal que se produce, dentro de la Unión Europea (UE), entre las innovaciones tecnológicas en materia de Inteligencia Artificial y sus regulaciones jurídicas. Para ello se parte de una tesis formulada por Hartmut Rosa, sociólogo alemán cercano a la Teoría Crítica de la sociedad, de acuerdo con la cual las estructuras temporales de la política no resultan hoy ya compatibles con el ritmo de cambio de algunas esferas sociales. Esto produce una nueva forma de patología social, que Rosa denomina “riesgos de la desincronización”, y cuyo efecto más preocupante hay que buscar en el desplazamiento de los procesos de toma de decisiones desde el ámbito de la política hacia otros ámbitos de la sociedad más rápidos, fundamentalmente el mercado. Para analizar esta problemática se reconstruye, en primer lugar, la tesis de la desincronización como patología social derivada de la aceleración. Tras ello, se hace un mapeo del proceso de legislación en materia de IA a nivel de la UE. Por último, se defiende la tesis de que este proceso legislativo, ciertamente muy admirable, ha llegado sistemáticamente tarde. (shrink)

¿Puede la inteligencia artificial (IA) hacernos más morales o ayudarnos a tomar decisiones más éticas? El libro Más (que) humanos. Biotecnología, inteligencia artificial y ética de la mejora, editado por Francisco Lara y Julian Savulescu (2021), puede inspirarnos filosóficamente sobre este debate contemporáneo. En esta nota crítica, contextualizo la aportación general del volumen y analizo los dos últimos capítulos de Monasterio-Astobiza y de Lara y Deckers, quienes argumentan a favor del uso de la IA para hacernos (...) mejores agentes morales. El objetivo es ampliar y matizar críticamente algunas cuestiones clave sobre cómo la IA puede asistirnos para tomar decisiones más éticas. (shrink)

The first part of this paper contextualizes the debate on the ethical regulation of Artificial Intelligence, and then reviews the main Anglo-Eurocentric elements of public initiatives about the ‘Ethics of AI’ that have recently emerged. All this with the ultimate purpose of analyzing the basis of principlism of the Ethics of AI. We refer to this with the specific purpose of presenting it in terms of a paradigmatic case of ‘colonized’ ethics that hides the different moral judgments and the (...) alternative cultural axiologies. We begin with the hypothesis that there is a possibility of questioning the real purpose and contribution of this applied ethics, using a dilemmatic perspective specific to human and legal sciences. In order to execute this type of analysis, which is highly required to understand what the future of these proposals is and how it can affect the regulation of AI, an analysis related to the contemporary literature from the Decoloniality field will be performed. (shrink)

Experts in Artificial Intelligence (AI) development predict that advances in the development of intelligent systems and agents will reshape vital areas in our society. Nevertheless, if such an advance is not made prudently and critically-reflexively, it can result in negative outcomes for humanity. For this reason, several researchers in the area have developed a robust, beneficial, and safe concept of AI for the preservation of humanity and the environment. Currently, several of the open problems in the field of AI (...) research arise from the difficulty of avoiding unwanted behaviors of intelligent agents and systems, and at the same time specifying what we really want such systems to do, especially when we look for the possibility of intelligent agents acting in several domains over the long term. It is of utmost importance that artificial intelligent agents have their values aligned with human values, given the fact that we cannot expect an AI to develop human moral values simply because of its intelligence, as discussed in the Orthogonality Thesis. Perhaps this difficulty comes from the way we are addressing the problem of expressing objectives, values, and ends, using representational cognitive methods. A solution to this problem would be the dynamic approach proposed by Dreyfus, whose phenomenological philosophy shows that the human experience of being-in-the-world in several aspects is not well represented by the symbolic or connectionist cognitive method, especially in regards to the question of learning values. A possible approach to this problem would be to use theoretical models such as SED (situated embodied dynamics) to address the values learning problem in AI. (shrink)

A Inteligência Artificial (AI) é a mais nova hype tecnológica. Embora envolva grande diversidade de tecnologias, uma imagem genérica, superpoderosa, fortemente negativa e, no limite, distópica da IA tem sido disseminada, especialmente por obras de ficção científica. Recentemente, sobretudo após o lançamento do ChatGPT, houve uma explosão de publicações jornalísticas e acadêmicas, que, em parte, reforçam esse imaginário. Nas abordagens acadêmicas, os riscos seguem em destaque e, inclusive, cenários apocalípticos relacionados à AI são levados em consideração. A preocupação ética, (...) então, assume um sentido eminentemente negativo e os possíveis benefícios das AIs, muito mais realistas, são subfocalizados. Neste artigo, destacamos a importância de ir além da hype, desenvolvendo abordagens multi/interdisciplinares, delimitadas, realistas e ponderadas. Recorrendo à inteligência coletiva expressa na interação entre pesquisadores de três áreas (ciência da computação, linguística e filosofia), abordamos uma ferramenta de tradução automática (Machine Translation) gratuita: Google Translate. Nosso objetivo é mostrar que abordagens bem delimitadas, multidisciplinares, tecnicamente respaldadas e que não adotem o discurso sensacionalista nos conduzem a reflexões epistemicamente mais consistentes e crítico-normativas mais ponderadas, o que seria fundamental para o avanço do debate sobre AI. (shrink)

This article investigates the metaphor of the “black box” in artificial intelligence, a representation that often suggests that AI is an unfathomable power, politically uncontrollable and shrouded in an aura of opacity. While the concept of the “black box” is legitimate and applicable in deep neural networks due to the in- herent complexity of the process, it has also become a generic pretext for the perception, which we seek to critically analyze, that AI systems are inscrutable and out of (...) control, as well as supposedly endowed with intel- ligence and creativity. To challenge these ideas, we will address what we call the supremacy of patterns and the two significant phenomena that result from it: enchanted determinism and the dictatorship of the past. (shrink)

Artificial intelligence is a general-purpose technology (GPT), term given to technologies that shape an entire era and reorient innovations by reconfiguring the economy’s logic and functioning and bringing in new business models. AI offers unprecedented opportunities and risks. The benefits of AI are extraordinary, as are its potential harms. Potential damage does not have the same degree of problematization, since the intensity and extent of the damage varies according to the domain and the object of application. To address the (...) scale of this challenge, regulation is necessary but not sufficient. Standards, unwritten codes of compliance and arbitration procedures, supervision and auditability, AI governance, international agreements, compliance with current local and global standards and laws. All of this needs to be integrated. Society seems to have no alternative to facing the challenges of at least mitigating the damage already identified and trying to predict future damage in advance. The purpose of this article is to encourage reflection regarding the main initiatives that are available to society to protect its citizens and organizations from the potential harm caused by AI models, vis-à-vis the technology’s own limits to act in ethical and legal compliance. (shrink)

La inteligencia artificial (IA) no es sólo una moda, sino un conjunto de sistemastecnológicos que han adquirido en los últimos años un enorme desarrollo con aplicaciones a múltiples aspectos de la vida cotidiana de los ciudadanos, tanto por parte de particulares (empresas) como de los propios Estados.La Unión Europea ha apostado siempre por compatibilizar los avances tecnológicos con el respeto de los principios y valores de la democracia, el Estado de Derecho y los derechos y libertades fundamentales. Desde (...) tales premisas, actualmente se enfrenta a la búsqueda de un modelo regulatorio de la IA ante las carencias del Reglamento General de Protección de Datos (RGPD) en este ámbito.A través de la Propuesta de Reglamento (UE) sobre IA, se pretende adoptar una regulación marco del uso de la IA que suponga una “tercera vía” inspirada en los principios y valores de la Unión y se contraponga a las actuales fórmulas regulatorias “abstencionistas” (EE.UU) e “hiperreguladoras” (China), proporcionando un modelo que consiga el efecto imitador o “efecto Bruselas” de una IA “confiable” para los ciudadanos y las empresas a nivel mundial. (shrink)

Finnegans Wake es la última obra escrita por el escritor irlandés James Joyce. En ella, la creciente complejidad narrativa que lo caracterizó alcanza niveles extremos, rozando lo ininteligible. El autor no sólo mezcla diferentes idiomas, también inventa nuevos términos que germinan y se significan, en gran medida, desde su dimensión sonora. Personajes, y acontecimientos reales y ficticios se mezclan, formando una amalgama donde variantes de arquetipos mitológicos transcurren una glocalidad cíclica perpetua. La protagonista femenina, Anna Livia Plurabelle (ALP), es un (...) personaje mutante que deviene diferentes elementos naturales a lo largo del texto. El murmullo del río se confunde con el suyo propio, en un pulso narrativo que sobrescribe la orilla en la historia. De fondo siempre se oyen voces y, aunque no son claras, sugieren. Esta obra literaria no se lee, se escucha; no se entiende, se deduce. La presente investigación se sitúa en el contexto actual de auge en la aplicación de la Inteligencia Artificial (IA) al procesado de textos. Somete a prueba la transcripción realizada a partir de la locución de algunos pasajes de Finnegans Wake, evaluando el grado de acierto y los posibles sesgos o desviaciones de los resultados. (shrink)

Resumo: neste artigo pretendo estender a evidenciação do encobrimento do mundo da vida formulada por Husserl ao uso das tecnologias disruptivas da 4ª Revolução Industrial, mais precisamente o uso da inteligência artificial (IA) nas plataformas sociais. Primeiro, apresento o conceito de mundo da vida que se mantém consistente ao longo das obras de Husserl; em seguida, apresento seu encobrimento derivado da matematização da natureza e do rompimento com o telos, o que originou a crise da humanidade europeia. Segundo, mais (...) camadas de encobrimento do mundo da vida vão se sobrepondo com a 3ª e 4ª RIs. Em mais detalhe, o encobrimento e o afastamento do mundo da vida provocado pelo uso da IA e seus efeitos negativos, individual e coletivamente. Terceiro, a proposta Husserliana de retorno ao mundo da vida através de uma epoché radical. Concluo que a solução proposta por Husserl para a crise ainda é atual e urgente. Abstract: in this paper I intend to extend the disclosure of the cover-up of the lifeworld formulated by Husserl to the use of disruptive technologies of the 4th Industrial Revolution, more precisely the use of artificial intelligence on social platforms. First, I present the concept of lifeworld that remains consistent throughout Husserl's works; then, I present its cover-up derived from the mathematization of nature and the break with the telos, which originated the crisis of european humanity. Second, more layers of cover-up of the lifeworld overlap with the 3rd and 4th IRs. In more detail, the cover-up and detachment from the lifeworld caused by the use of AI and its negative effects, individually and collectively. Third, the Husserlian proposal of return to the lifeworld through a radical epoché. I conclude that the solution proposed by Husserl to the crisis is still actual and urgent. (shrink)

This article is intended to present a proposal for dialogic digital ethics on a critical reading of the European Commission's independent high-level expert group’s document Ethics Guidelines for Trustworthy AI (2019). These would be digital ethics with a normative horizon for action and criteria for justice based on dialogue and possible agreement between all agents involved and affected by the digital reality. The aim is to show that the participation of all parties involved is not merely advisable but morally required (...) as part of the Commission’s effort to generate common European willingness and governance to deal with the fourth industrial revolution now going on. The acknowledgement of equal dignity implied by people-centric artificial intelligence (AI) is utterly unthinkable without this possibility of equal participation. Without it, trust cannot be generated or guaranteed. As we intend to show, the new principle of explicability plays a decisive role in this objective as a principle with a moral as well as an instrumental. El presente artículo tiene como objetivo presentar los rasgos básicos de una ética digital dialógica a partir de una lectura crítica del documento elaborado por grupo independiente de expertos de alto nivel para la Comisión Europea Ethics Guidelines for Trustworthy AI (High-level expert Group on Artificial Intelligence, 2019). Una ética digital que tiene en el diálogo y acuerdo posible de todos los agentes implicados y afectados por la realidad digital su horizonte normativo de actuación, su criterio de justicia. La finalidad es mostrar que, en su esfuerzo por generar una voluntad común y una gobernanza europeas ante la actual revolución industrial, la participación de todas las partes implicadas no solo es recomendable sino moralmente exigible. El reconocimiento de la igual dignidad que implica una Inteligencia Artificial centrada en las personas no es ni siquiera pensable sin el horizonte de una participación igual. Sin ella, la confianza no puede generarse ni garantizarse. Como pretendemos mostrar, en este objetivo juega un papel decisivo el nuevo principio de explicabilidad, principio que posee un valor moral y no solo instrumental. (shrink)

Artificial intelligence (AI) techniques are used to model human activities and predict behavior. Such systems have shown race, gender and other kinds of bias, which are typically understood as technical problems. Here we try to show that: 1) to get rid of such biases, we need a system that can understand the structure of human activities and;2) to create such a system, we need to solve foundational problems of AI, such as the common-sense problem. Additionally, when informational platforms uses (...) these models to mediate interactions with their users, which is a commonplace nowadays, there is an illusion of progress, for what is an increasingly higher influence over our own behavior is took for an increasingly higher predictive accuracy. Given this, we argue that the bias problem is deeply connected to non-technical issues that must be discussed in public spaces. -/- Técnicas de inteligência artificial (IA) são utilizadas para modelar as atividades humanas e gerar predições comportamentais. Estes sistemas têm apresentado vieses diversos, inclusive de raça e gênero, tipicamente tomados como problemas de engenharia. Realiza-se aqui um esforço argumentativo para mostrar que: 1) escapar dos vieses demanda um sistema que compreenda a estrutura das atividades humanas e; 2) criar um sistema que apresente tal compreensão, demanda a solução de problemas fundacionais da IA, em particular, o problema de como modelar o senso comum. No caso de plataformas informacionais que usam desses modelos para mediar interações com seus usuários, algo cada vez mais comum, o não reconhecimento disso dá margem a uma ilusão de progresso, em que uma crescente influência sobre nosso comportamento é tomada como uma crescente acurácia preditiva. Nesse cenário, argumenta-se que o problema dos vieses está associado a questões não técnicas que devem ser discutidas em espaços públicos. (shrink)

The purpose of this paper is to study the philosophical basis of the architecture of two Artificial Intelligence systems (AlphaGo and Hide and Seek). The problem considers the elucidation of the epistemological distinction between “knowledge” and “intuition” and questions whether the fulfillment of a programmed metric can be considered as a product of knowledge or, as an example of intuitive thinking. Through the analysis of both examples, it is shown that achieving the desired objective (winning a game of Go (...) and developing new hiding strategies) is not derived from the system’s architecture, but from the AI’s creativity to achieve the desired result; in that sense, it can be confirmed that certain intelligent systems are capable of creative and intuitive thinking. El objetivo del presente artículo es estu- diar los fundamentos filosóficos de la arquitectura de programación en dos sistemas de Inteligencia Artificial (AlphaGo y Hide and Seek). El problema dilucida la distinción epistemológica de los concep- tos “conocimiento”, “intuición” y “abducción”, para definir si el cumplimiento exitoso de una métrica programada, por parte de una computadora, es con- dición suficiente para atribuirle un comportamiento inteligente. A través del análisis de ambos ejemplos se muestran dos cuestiones: i) el cumplimiento exitoso de un objetivo programado ofrece nuevos recursos de conocimiento, ii) dichos conocimientos dependen de la ejecución de un programa cuyo pro- cesamiento es desarrollado por una IA y, por tanto, las operaciones superan las capacidades intelectivas humanas. Las conclusiones apuntan a que las com- putadoras son recursos de conocimiento especiales de comprobación de hipótesis. (shrink)

Artificial intelligence in all its developments is advancing faster than the capacity of institutions and organizations to offer legal, but also deontological responses: assuming the ethical implications of the new digital scenarios where technology learns from human routines to personalize content is fundamental from the academic, politicy and business spheres. This paper offers an interpretative framework of the evolution of smart speakers and voice assistants as tools for information singularization in parallel to the study of the European framework for (...) the creation of an ethics of ai. The results show an effort by institutions to respond to an unstoppable progress and demonstrate that ai cannot be addressed by legal regulation alone but is above all a matter of ethical preference. (shrink)

Within the strong branch of artificial intelligence, which is aimed at creating thinking machines with intellectual powers like those of man, the most explored research program is symbolic aI, defined as the attempt to use electronic computers to replicate the human mind, either assuming a structural and functional similarity between them, or trying to replicate the behavior produced by the human mind through computational processes that also have an intentional structure but are only instrumentally equivalent. In this paper we (...) show that strong symbolic aI, in either of its two variants, is impossible, since formal systems, that is what all computer programs ultimately are, are not sufficient to replicate in an intentional way four intellectual faculties of man that are essential to intelligent behavior: sense of the situation, common sense, procedural skills and theory abduction. (shrink)

The central goal of this paper is to investigate the injustices that can occur with the use of new technologies, especially Artificial Intelligence (AI), focusing on the issues concerning respect to human rights and the protection of victims and the most vulnerable. We aim to study the impacts of AI in daily life and the possible threats to human dignity imposed by it, such as discrimination based on prejudices, identity-oriented stereotypes, and unequal access to health services. We characterize such (...) cases as algorithmic injustices. In the final part of the text, we propose some strategies to confront this specific type of injustice. (shrink)