Absence of information in decision-making processes (ABSENCE)

The project examines the differences of human and machine decision-making processes through philosophical investigations of the role absence plays in these processes. Absence seems to be a productive force for human communication (omission, insinuation) and decision-making, but how does it influence algorithmic systems?

We propose that the investigations of absence in human decision-making processes will further our understanding of artificial intelligence and its limitations and will be especially beneficial in contexts of automated decision-making. The focus on absence of information enables us to address ‘that which is present without being present’ in algorithms in ways that move beyond current ethical concerns regarding discrimination, bias, and unfairness. To these ends we propose the metaphor of Algorithmic Implicature as a way to explore and understand that which is present without being explicit in the automated domain – i.e. implicit attitudes, intentions, and assumptions. The grand question lying underneath is what it means to be human in relation to ‘intelligent’ machines.

Research focus

The project will conceptualize absence (of information) as a productive and inevitable force in algorithmic processes for decision-making by exploring human social intelligence.

Objective and research questions

Exploring and delimiting algorithmic implicature
How does absence of information condition social intelligence as manifested in communication?
How does absence-based inference play into decision-making?

Approach

The project is situated within philosophy of information, an interdisciplinary field between philosophy, information studies, and computer science, including information ethics, privacy studies, surveillance studies, and philosophy of technology. Further, the project takes a language philosophic and epistemological approach inspired by recent developments in philosophy of technology focusing on conceptual analyses complemented by interviews with experts working with AI in order to explore what is possible when dealing with absence.

Subprojects

How does absence of information condition social intelligence as manifested in communication?: Our objective is to explore further the pragmatic realm of human communication in relation to machines – how absence plays into communication. It speaks to the human ability of abstract thinking and concept formation. In this part of the project, the concept of absence takes center stage. How does absence of information work in communication? How can we conceptualize it?
How does absence-based inference play into decision-making?: Given that human communication is replete with absence and we never have full information awareness, it is crucial to explore absence-based inference. We will advance the theoretical and philosophical work on absence-based inference – a very fundamental but undertheorized concept. The philosophical advancements will inform the metaphor of algorithmic implicature and will – together with expert interviews – inform and advance our understanding of automated decision-making.