Inteligencia Artificial y sesgos de género

• Alonso Betanzos, Amparo ^[1]
  1. [1] Universidade da Coruña

     Universidade da Coruña

     A Coruña, España

     
• Localización: Gender on digital: Journal of Digital Feminism, ISSN-e 3020-4410, V. 1, 2023, págs. 11-32
• Idioma: español
• DOI: 10.35869/god.v1i.5060
• Títulos paralelos:
  • Artificial intelligence and gender bias
• Enlaces
  • Texto completo
• Resumen
  • español

    Estamos inmersos en una nueva revolución, una era de transformación impulsada por la Inteligencia Artificial (IA), que afecta significativamente al equilibrio geopolítico, la sociedad, la economía, el empleo y la educación, generando cambios constantes en estos ámbitos. La IA es una disciplina transversal que está presente en prácticamente cualquier campo, desde la Industria, la Salud o el Medioambiente hasta áreas relacionadas con las Ciencias Sociales y las Humanidades. Esta omnipresencia trae consigo innumerables oportunidades y abre nuevas perspectivas, pero también trae retos, algunos relacionados con ciertos desafíos éticos que pueden aparecer por el tratamiento de datos a gran escala que hace la tecnología. Uno de ellos es la posible aparición de sesgos de género, que pueden deberse a que el funcionamiento de los algoritmos no ha sido suficientemente examinado en este sentido, o a que el entrenamiento de los modelos se ha realizado con datos históricos cuya calidad no es la adecuada, entre otros. Además, es fundamental tener en cuenta que los sesgos no sólo están en los algoritmos, también pueden derivar de desigualdades en el acceso a la tecnología, o de falta de diversidad en los equipos de diseño, entre otros. Estos factores pueden limitar la perspectiva y comprensión holística de los problemas, perpetuando así prejuicios y desigualdades en la IA. Es imprescindible abordar con responsabilidad estos sesgos si queremos garantizar una IA ética, confiable y justa. No es menos importante fomentar la diversidad en los equipos de desarrollo y diseño de tecnologías de IA, para que se reflejen diferentes perspectivas que puedan conducir a soluciones más inclusivas y equitativas.

  • English

    We are immersed in a new revolution, an era of transformation driven by Artificial Intelligence (AI), which significantly affects the geopoliticalbalance, societal norms and behaviour, the economy, employment, and education, generating constant changes in these areas. AI is a transversal discipline that is present in practically any field, from Industry, Health, or the Environment to areas related to the Social Sciences and Humanities. This omnipresence brings with it innumerable opportunities and opens new perspectives, but it also brings challenges, some related to certain ethical issues that can arise from the large-scale processing of data of AI algorithms. One of them is the possible appearance of gender biases, which may be because the operation of the algorithms has not been sufficiently examined in this sense, or that the training of the models has been carried out with historical data whose quality is not adequate, among other things. In addition, it is essential to consider that biases can also derive frominequalities in access to technology, or from a lack of diversity in design teams. These factors can limit the holistic perspective and understanding of the problems, thus perpetuating prejudices and inequities in AI. It is essential to responsibly address these biases to guarantee ethical, trustworthy and fair AI. It is no less important to encourage diversity in the development and design teams of AI technologies so that different perspectives are reflected which can lead to more inclusive and equitable solutions.

• Referencias bibliográficas
  • Chat GPT. OpenAI, https://chat.openai.com
  • González de la Garza, Luis M. Teoría de sesgos en el sistema educativo de la democracia del siglo XXI Nuevas garantías para...
  • Regulation 2016/679, General Data Protection Regulation, https://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:32016R0679.
  • S. Leavy, "Gender Bias in Artificial Intelligence: The Need for Diversity and Gender Theory in Machine Learning,"2018 IEEE/ACM...
  • High-Level Expert Group on Artificial Intelligence. Ethical Guidelines for Trustworthy AI. April, 2019.
  • Proposal for a regulation of the European Parliament and of the Council laying down harmonized rules on artificial intelligence...
  • European Commission, Directorate-General for Justice and Consumers, Gerards, J., Xenidis, R., Algorithmic discrimination in Europe...
  • Turing, Alan M. (1950) Computing Machinery and Intelligence. Mind 49: 433-460. https://redirect.cs.umbc.edu/courses/471/papers/turing.pdf
  • https://digital-strategy.ec.europa.eu/en/policies/desi
  • West, Mark, Kraut, Rebecca, EiChew Han. I’d blush if I could: Closing gender divides in digital skills through education. UNESCO...
  • Blumberg,Sven, Krawina, Melanie, Mäkelä, Elina and Soller, Henning. Women in tech: The best bet to solve Europe’s talent...
  • Torre de la Sierra, Ana M., and Guichot-Reina, Virginia. The influence of school textbooks on the configuration of gender identity:...
  • Sunderland, J. New understandings of gender and language classroom research: Texts, teacher talk and student talk Language...
  • Obermeyer, Ziad; Powers, Brian; Vogeli, Christine and Mullainathan, Sendhil. Dissecting racial bias in an algorithm used to manage...
  • Turner Lee, Nicol; Resnik, Paul and Barton, Genie. https://www.brookings.edu/articles/algorithmic-bias-detection-and-mitigation-best-practices-and-policies-to-reduce-consumer-harms/,...
  • Martin, Kirsten. Ethics of Data and Analytics, Auerbach Pub, 2022.
  • Merill, Jeremy B. https://themarkup.org/google-the-giant/2021/02/11/google-has-been-allowing-advertisers-to-exclude-nonbinary-people-from-seeing-job-ads,...
  • Bolukbasi, Tolga; Chang, Kai-Wei; Zou, James; Saligrama, Venkatesh and Kalai, Adam . Man is to Computer programmer as Woman is to Homemaker?...
  • Raji, Inioluwa Deborah; Gebru, Timnit; Mitchell, Margaret; Buolamwini, Joy; Lee, Joonseok, and Denton, Emily. Saving face: Investigating...
  • Document 52022PC0068. Proposal for a REGULATION OF THE EUROPEAN PARLIAMENT AND OF THE COUNCIL on harmonized rules on fair access...
  • European Commission, ‘White Paper on Artificial Intelligence: A European approach to excellence and trust’ COM(2020) 65 final, 2020.
  • European Commission. Communication from the Commission to the European Parliament, the Council, the European Economic ...
  • Gerards, Janneke and Xenidis, Raphaëlle. Algorithmic discrimination in Europe: Challenges and opportunities for gender ...