Program

Algorithmic bias and polarization constitute an ever growing danger to our online media landscape. The consequences affect not only individuals themselves – who are often locked in the filter bubbles of their personal news feeds –, but also results in severe societal damage. The goal of this presentation is to show, suggest and discuss methods for quantifying different forms of bias – wanted and unwanted, natural and algorithmic – and to reduce the unwanted effects of such bias, while maintaining the benefits of a pluriform media landscape. During the presentation, I will show results from an analysis on perceived bias in the Chilean media, based on a recently published journal article.

In this talk I introduce epistemic justifications of democracy, show how interactions on social networks might undermine central premises of such justifications, and ask whether a tighter regulation of social media platforms can be justified. One might think that the right to a free press and media ties the hands of regulators. I will aim to show that this view is mistaken. By unpacking the core values underpinning media freedom, I show that setting limits to media freedom may be required to uphold the core values that underpin the normative ideal of free media.

In this presentation, I am going to discuss a series of diverse interdisciplinary research projects that I have been working on the last three years, which I believe can contribute substantially to the development of a project on machine ethics issues in modern journalism. More specifically, I am going to elaborate on papers I have published and others I am currently developing on the topics of: i) evolution of social networks, ii) online/offline interactions, iii) user behaviour on online and virtual spaces, as well as iv) privacy, rights and security.

In light of the current explosion of application of machine learning in data
analysis and inference, we examine a particular challenge raised by the new
EU General Data Protection Regulation (GDPR). The challenge we address
pertains particularly to the demand that analyses of a person’s data must be
comprehensible to that person.
While there is a long tradition in viewing the world in terms of objects
and properties in intuitive ways, recent decades have entertained a tension
between more rule-based theories of mind (e.g., the representational theory
of mind) and more holistic approaches (e.g., connectionism). While both
approaches have merit, one seems to depart too much from a classical
understanding of “knowing” to adequately satisfy the imminent legal realisty,
and the other seems to be incapable of adequately capturing modern data
analysis (as of yet).
As a solution to this predicament we propose a pragmatic compromise
based on argumentation theory which seems to be able to provide a
solid foundation in classical concepts, while at the same time permitting
enthymematic presuppositions. We argue that developing a framework for
explaining machine behavior in terms of abstract argumentation theory can
address this dilemma.