The article looks at the mechanisms of Artificial Intelligence related to the processing of human knowledge and natural language from a multilingual perspective. Undoubtedly, AI has made very impressing progresses in these areas, but they are unsatisfactory when it comes to what is called “the long tail”, i. e. the interpretation of less frequent words or concepts. In addition, AI has deficiencies when the context plays an important role, which is often the case. Most of the knowledge resources and methods actually used by AI have not been modelled to take multilingual and multicultural aspects into consideration. The article describes these issues and suggests some remedies, opening new opportunities for translators and interpreters.
© 2001-2024 Fundación Dialnet · Todos los derechos reservados