Paper short abstract:

Shaping social realities of people in digitized environments, AI technology relies on models of biological information processing. This paper explores biological factors omitted in technological abstraction and their potential changes to AI embodiment in a combined bioethics and AI ethics framing.

Paper long abstract:

Based on mathematical models of biological learning processes, computational algorithms are forming the basis of machine learning or artifical intelligence (AI). Following a largely dualistic and disembodied understanding of human mental processes, their technological abstraction is creating a wealth of application opportunities and promises immense transformative potential across social sectors.

AI ethics discuss effects and desirability of such transformations for living realities of affected people; however, a merely technology-centered discourse about desirable social effects of widespread AI implementation overlooks the technology’s origins in biological information processing from a life sciences and human sciences perspective as well as its complex, multi-layered effects on psychic, social and cultural systems as humanly embodied entities.

Especially with the rise of emerging neurotechnologies enabling novel interfaces between biological and technologically abstracted forms of intelligence, acknowledging these aspects in capable and integrated ethical concepts is gaining increasing importance. Drawing on central concepts in bioethics and AI ethics, an integrated embodiment approach can strengthen these reflexive elements in AI ethical and regulatory debates, as well as improve social discourse and agency relating to technology-driven transformations of living environments as application fields of AI technology.

Biological factors other than electrophysiological activity (e.g. chemical, hormonal, metabolic signals) are strongly influencing neural or bacterial information processing, yet they are omitted in AI technological abstraction. This paper explores technoscientific imaginaries potentially arising from their inclusion in AI development, and how social interaction with the differently embodied technology might change if more diverse sources of bioinformation were included in AI production.