Does A.I. Speak Our Language? A Gadamerian Assessment of Generative Language Models

• In Event: Artificial Intelligence: Concerns for Politics and for the Profession

Abstract

The language argument is a classic argument for human distinctiveness that has separated humans from other non-human animals for millenia. Only very recently has the possibility of AI employing something like natural language become a reality. The explosion of public and scholarly interest in generative language models in the last year (e.g., OpenAI’s ChatGPT, Google’s Bard, or Microsoft’s Bing) is evidence of this. The generative language models are impressive for being able to construct coherent, grammatically correct, and informative prose. For this reason, generative language models pose a challenge to traditional language-based models of human distinctiveness precisely because they can communicate and respond in a manner resembling humanity’s linguistic capabilities. This paper asks: have generative language models finally acquired natural language? Employing Gadamer’s theory of language, I argue that they have not. While a language AI can reliably generate linguistic content that can be interpreted as “texts,” generative language models lack a linguistic construction of reality. Missing from these models are four key features of a linguistic construction of reality: groundedness to the world, understanding, community, and tradition. I conclude with skepticism that AI models can ever achieve natural language because they lack these characteristics in their linguistic development.

Author

• Phillip Marcial Pinell, University of Wisconsin-Madison