El próximo martes, 29 de octubre, a las 14.30, nuestro compañero Ivar Rodríguez Hannikainen, investigador de FiloLab, dará una charla-seminario en el Max Planck Institute for Human Development de Berlín, invitado por el Center for Humans and Machines. A continuación se reproduce el título y un resumen de su charla, que se podrá seguir online a través de Webex. Podéis encontrar los detalles en este enlace.
Title: Authorial credit and trust in LLMs: artifice or estrangement?
Abstract: In this talk, I will present ongoing work examining human interactions with LLMs in the domains of informational content creation, artistic production, and moral counseling. How do interactions with LLMs influence credit for co-created outputs? And why do humans deposit or withdraw trust in LLMs? I tackle these questions with two hypotheses in mind: (i) the artifice hypothesis, according to which the artificial architecture of LLMs inflexibly undermines credit and trust, and (ii) the estrangement hypothesis, according to which barriers to credit and trust can be surmounted by personalization/appropriation.
