Josh Martin

Este artículo analiza cómo la inteligencia artificial reproduce y consolida la normatividad androcéntrica a través de las affordances que surgen en el diseño, el entrenamiento y el uso de los sistemas algorítmicos. Basándonos en la teoría de las affordances, que entiende estas como invitaciones a la acción, sostenemos que los artefactos tecnológicos no solo facilitan determinadas prácticas, sino que también encarnan normas implícitas, dando forma a modos de acción situados culturalmente. Aunque la bibliografía sobre el sesgo algorítmico ha documentado manifestaciones de desigualdad de género, a menudo carece de un marco que explique cómo dichos sesgos se consolidan como normas a través de la interacción entre el usuario y el sistema. Para subsanar esta laguna, proponemos una tipología de affordances normativas androcéntricas en tres niveles: (i) datos y corpus de entrenamiento, (ii) arquitecturas y objetivos de diseño, y (iii) configuraciones de interfaz y prácticas de uso. Esta tipología se ilustra con minicasos, como la traducción automática de Google, que históricamente ha asignado las profesiones de cuidados a las mujeres. Al conceptualizar el sesgo androcéntrico como un conjunto de affordances normativas, el artículo ofrece un marco crítico para diagnosticar cómo los sistemas de IA normalizan las jerarquías de género y para replantear su diseño hacia prácticas sociotécnicas más democráticas.

Universal Credit, the UK working-age benefit, attempts to address issues of complexity, inflexibility, and work disincentivisation within the benefits system. However, the UK coalition government’s delayed rollout and emphasis on conditionality raise questions over its applicability to twenty-first century labour market problems. Instead, universal basic income tackles similar goals and arguably better addresses labour market changes. But is this transition politically feasible? Through interviews with policy actors from four UK political parties, arguments for and against the likelihood of this transition are presented by utilising theories of welfare state development.

The book proposes an alternative model of technological development, that of engaging technologies. This model studies the way in which technological development can be perfect towards technological solutions compatible with sustainable development, with the ideals of solidarity or with cooperative and or only competitive participation. In short, the book is an analytical, descriptive and normative journey around the concept of "endearing technologies" coined by Miguel Ángel Quintanilla, which highlights the democratic values for the control of technology as opposed to the purely intrinsic values of technical efficiency and performance, as opposed to the automatic and uncontrolled tendency of technology.

Ignorance is often a valid excuse for wrongdoing. But authors such as William FitzPatrick argued that ignorance is culpable if we could have reasonably expected the agent to take action that would have corrected or prevented it, given his capabilities and the opportunities provided by the context, but failed to do so due to vices such as laziness, indifference, disdain, etc. Guilty ignorance is still present in the debate and, in recent times, has become more pressing with the problem of technological responsibility. In this paper, an internalist perspective of culpable ignorance is adopted to analyze a form of culpability distribution in the technological context based on dispositional beliefs. Thus, two types of responsibility are found. By examining the implications of culpable ignorance, we realize that we can respond to the unambiguous idea that an engineer is morally and epistemically responsible for certain facts.