This is an introductory post to my master’s thesis. Tldr; I am conducting a content analysis on all AI-related policy papers and speeches the Singapore government has published in order to understand the state of AI governance in Singapore. The interactive visualisation can be accessed here.

Background

Everyone—governments, private companies, the public—is engaging in AI governance discourse. And because everyone has an own opinion on how AI will change the world (for better or for worse), and how it should be progressed and regulated, discourse seeks to reach a happy ending (consensus) on what norms, policies and institutions determines the path of AI’s future.

With all the different priorities, actors, and sentiment, AI governance discourse is messy, with conflicting definitions, interpretations and methods of implementation. But that just reflects the complexity of our social reality. Plurality is all part of healthy discourse.

Yet, we find that the literature on AI governance discourse tips in favour to developments in the West, with a lack of engagement with its eastern complement. This is despite there being plenty of evidence that nations in Asia are also increasingly partaking in AI governance discourse, presenting rich grounds for empirical work.

Site

My thesis aims to contribute to the body of AI governance discourse by constructing a conceptual map of AI in Singapore. The launch of the Smart Nation Initative in 2014 could be regarded as the beginning of the nation’s engagement in AI governance discourse. Followed up by a National AI Strategy in 2019 and topping the Smart City Index in 2020, Singapore’s active engagement in the application and regulation of AI makes it a rich and grounded case study for AI governance discourse in Asia.

Methodology

I approach constructing this conceptual map through a content analysis of policy papers and speeches by the government on the topic of AI. All the data is publicly accessible through the various ministry websites. The content analysis aims to identify the descriptors of AI systems, the issues to which AI is being applied, as well as actors involved.

I will be take a computational approach, in both constructing a coding dictionary, as well as applying natural language processing (NLP) techniques—topic modelling, named entity recognition, etc.

Aim

Uncovering the perspective taken towards AI systems by the government is important as how a government perceives AI technology shapes the interpretation of its capabilities and potential, and consequently, the policies enacted to govern and regulate the development and application of AI systems. Having a clear picture of the state of AI governance in Singapore will allow us to identify its gaps, conflicts and ethical conundrums.

The conceptual map I aim to build can also be constructed as a map of representations, revealing the perspective towards AI systems that is being communicated by the government to the recipients of these policy documents and allowing us to potentially identify framing strategies and agenda setting around AI governance.