Artificial agency and the game of semantic extension

ABSTRACT

Artificial agents are commonly described by using words that traditionally belong to the semantic field of life. I call this phenomenon the game of semantic extension. However, the semantic extension of words as crucial as ‘autonomous’, ‘intelligent’, ‘creative’, ‘moral’, and so on, is often perceived as unsatisfactory, which is signalled with the extensive use of inverted commas or other syntactical cues. Such practice, in turn, has provoked harsh criticism that usually refers back to the literal meaning of the words to show their inappropriateness in describing artificial agents. Hence the question: how can we choose our words appropriately and wisely while making sense of artificial agents? This paper tries to answer by sketching the main features of the game of semantic extension in relation to artificial agency, reviewing the related opportunities and risks, and advancing some practical suggestions on how to play the game well.

KEYWORDS:

• Artificial agency
• language game
• game of semantic extension
• homology
• analogy
• similarity
• difference
• philosophy of technology

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 Many examples can be found, for instance, in the debate on trust and digital technologies: ‘e-trust’ in (Taddeo and Floridi 2011), ‘TRUST’ in (Grodzinsky, Miller, and Wolf 2011), ‘robotrust’ in (Pagallo 2010).

2 See, for example, (Coeckelbergh 2017b, 296): ‘if more people were to speak about what machines do in terms of ‘artistic creations’ and ‘works of art’, than would we really have an objective basis for saying that they are wrong? Even if today we might be opposed to the very idea of machine art, in the course of time, our language might change and let the machines in through the backdoor’.

3 In order to do so, I extensively draw on the book I just quoted that has already proven extremely useful in this inquiry, i.e. Wittgenstein’s Philosophical Investigations (Wittgenstein 1958). Other important observations on what it means to play a game and what this entails in relation to meaning can be read in Hans-Georg Gadamer’s Truth and Method (Gadamer 2004, 102–130).

4 For an example, see (Fossa 2017).

5 See, for plenty of examples, the use of proxies for measuring human virtues such as loyalty or dependability and the connected epistemological and ethical issues brought up by (O’Neill 2016). From the point of view of biology, see (Boldt 2018). This issue is also extremely visible in Wiener’s theoretical writings on cybernetics.

6 For an example, see (Fossa 2018).

7 I tried to do this in (Fossa 2019).

Additional information

Notes on contributors

Fabio Fossa

Fabio Fossa (Ph.D., University of Pisa) is a researcher at the Department of Mechanical Engineering of the Politecnico of Milan, Italy. His main research areas are applied ethics, philosophy of technology, robot and AI ethics, and the philosophy of Hans Jonas. His current research deals with the philosophy of artificial agency and the ethics of autonomous driving.