The article exposes how one of the most fundamental oppositions in the history of modern philosophy – the opposition between Hume’s empiricism and Kant’s critical philosophy – is embedded in the current debate on the impact of artificial intelligence (in particular, the algorithmic selection of content) on human society. Hume’s empiricism – with its deduction of subjectivity based on a process of habituation – corresponds to the functioning of recommending algorithms, while Kant’s idea of autonomous subjectivity corresponds to the ideals underlying today’s ethical attempts towards the regulation of artificial intelligence. According to such ethics, the use of empirical data can endanger humans; whereby our attention can be easily caught by sensationalist content and our autonomy replaced by the agency of machinic intelligence. However, as argued in the present article, such ethical positioning also reproduces the gap between the empirical reality and normative principles, which is why transcendental (Kantian) ethics should be supplemented with Hume’s immanent and practical reasoning.
Read full abstract