Study: Human-Centered AI Talks About People, But Does Not Define Them

Human-centered AI has emerged as one of the key slogans when discussing the impacts of AI on society. In a new article published in the AI & SOCIETY journal, researchers examine what the word 'human' actually means when talking about human-centered AI.

The study analyzes key literature on human-centered AI using so-called critical hermeneutics. Hermeneutics refers to the interpretation of texts and the uncovering of underlying assumptions. The aim is to determine what kind of image of humanity is hidden when talking about AI that is 'for humans' or built 'around humans.'

The authors identify five features that repeatedly shape the conception of humans in discussions about human-centered AI: universalism, individualism, instrumentalism, psychologism, and exceptionalism. Universalism assumes a single, universally applicable conception of humanity, individualism emphasizes the individual over communities and structures, instrumentalism primarily sees humans as users of tools, psychologism emphasizes the human mind at the expense of other dimensions, and exceptionalism views humans as a distinct species fundamentally separate from the rest of nature.

The article also introduces the thoughts of philosopher Hannah Arendt on humanity into the discussion. Arendt's so-called philosophical anthropology offers a way to understand human existence that, according to the study, differs from the assumptions of the current human-centered AI discourse. The researchers suggest that Arendt's perspective can help clarify what 'human' means when we create rules, systems, and ethical principles for AI.

Source: Making sense of the ‘Human’ in human-centered AI: an Arendtian perspective, AI & SOCIETY.