11. “Action” and Ascription: On Misleading Metaphors in the Debate About Artificial Intelligence and Transhumanism

In all areas of science communication, metaphors are a very important tool for explaining complex matters in everyday language. While in mathematics or physics, the domain language is visible as such, the notions used in the interdisciplinary field of artificial intelligence (AI) already seem like everyday language and therefore hide their inner complexity. AI vocabulary comprises many anthropomorphisms to describe seemingly cognitive-like computer functions. Considering terms such as “learning”, “deciding”, “recognizing” or even “intelligence” itself, purely descriptive conversations about the field are nearly impossible. Difficulties now arise when such charged vocabulary as in the field of AI is being directly transferred into other domains or back into everyday language used in political or public debates. Confusion expands even more, when such assumed capabilities hit the discourse concerning transhumanism, who wants to “enhance” humans technologically. However, “enhancement” is usually meant in a masculine and capitalist regime of enhancement: think faster, jump higher, live longer, be more productive.

This chapter makes the claim that present debates around AI and transhumanism are characterized by a general lack of sensitivity and critical distance to the various levels and contexts of metaphors used. As a consequence, questions regarding responsibility and potential use cases are greatly distorted. The chapter proves its claim by critically analysing core notions of the AI and transhumanism debates and is a conceptual work at the intersection of computer science/machine learning and the philosophy of language.