Just what are we doing when we're describing AI? Harvey Sacks, the commentator machine, and the descriptive politics of the new artificial intelligence

Mair, Michael ORCID: 0000-0003-0929-5426, Brooker, Phillip ORCID: 0000-0003-1189-4647, Dutton, William and Sormani, Philippe (2021) Just what are we doing when we're describing AI? Harvey Sacks, the commentator machine, and the descriptive politics of the new artificial intelligence. QUALITATIVE RESEARCH, 21 (3). pp. 341-359.

Access the full-text of this item by clicking on the Open Access link.

	Text Mair, Brooker, Dutton, Sormani - Harvey Sacks, the Commentator Machine, and the Descriptive Politics of the New Artificial Intelligence.docx - Author Accepted Manuscript Available under License : See the attached licence file. Download (411kB)
	Text Mair & Sharrock - Action, Meaning and Understanding, Seeing Sociologically with Harvey Sacks, author accepted manuscript.docx - Author Accepted Manuscript Available under License : See the attached licence file. Download (644kB)
	Text Mair, Brooker, Dutton & Sormani - Just What Are We Doing When We're Describing AI.pdf - Published version Available under License : See the attached licence file. Download (283kB) | Preview

Abstract

<jats:p> In dialogue with the work of Heather Love and colleagues, this article makes use of a peculiar ‘descriptive assemblage’ proposed by Harvey Sacks (1963) – that of the ‘commentator machine’ – to open up issues of ‘descriptive politics’ in the field of contemporary Artificial Intelligence (AI). We do so by reviewing the gameplay of Google DeepMind’s AlphaGo – an algorithm designed to outperform human players at the game of Go – with a focus on the incongruities of the much discussed, indeed (in)famous ‘move 37’ in a human-versus-machine challenge match in 2016 (e.g. Silver et al., 2017). Looking at move 37 in conjunction with the various layers of commentary that came to be woven around it, we explore the kinds of descriptive work involved in characterising the move, the troubles that work reveals and what we can learn about the practices and politics of description from encounters with ‘New AI’ applications like AlphaGo. </jats:p>

Item Type:	Article
Uncontrolled Keywords:	Ethnomethodology, conversation analysis, description, politics of method, Heather love, Harvey Sacks, artificial intelligence, STS, AlphaGo, the commentator machine
Divisions:	Faculty of Humanities and Social Sciences > School of Law and Social Justice
Depositing User:	Symplectic Admin
Date Deposited:	02 Mar 2022 10:10
Last Modified:	18 Jan 2023 21:11
DOI:	10.1177/1468794120975988
Open Access URL:	https://journals.sagepub.com/doi/pdf/10.1177/14687...
Related URLs:	Author Publisher
URI:	https://livrepository.liverpool.ac.uk/id/eprint/3149530

Available Versions of this Item

• Just What Are We Doing When We’re Describing AI? Harvey Sacks, the Commentator Machine, and the Descriptive Politics of the New Artificial Intelligence. (deposited 06 Oct 2020 15:34)
  • Just what are we doing when we’re describing AI? Harvey Sacks, the commentator machine, and the descriptive politics of the new artificial intelligence. (deposited 07 Dec 2020 09:55)
    • Just what are we doing when we're describing AI? Harvey Sacks, the commentator machine, and the descriptive politics of the new artificial intelligence. (deposited 02 Mar 2022 10:10) [Currently Displayed]