The impact of artificial intelligence – an economic analysis (AN 24/06)

Executive summary #

This paper introduces economic frameworks for assessing the impact of artificial intelligence (AI). It then provides a qualitative assessment of the implications of AI for New Zealand, given what we know about the underlying economic trends here. This note is intended to be a conversation starter, highlighting key areas that could be explored in more detail.

The paper addresses three issues, summarised as follows:

• Impacts of AI on productivity and investment
  Our understanding of the impacts of AI on productivity and investment is emerging as the technology develops. AI can be seen as a “General Purpose Technology”, that slowly diffuses across a range of sectors, but it may also serve as an “Invention of a Method of Invention”, transforming innovation across a wide range of fields. However, New Zealand’s traditionally slow diffusion of new technology and low levels of investment in intangible capital could be a barrier to realising AI’s benefits.
• Impacts of AI on employment and the labour market
  The net impact of AI on employment and the labour market, balancing job destruction (“displacement”) with creation (“reinstatement”) remains uncertain. AI’s disproportionate impact on higher-skilled tasks might mean employment in higher-skilled, advanced economies like New Zealand is more exposed to the impacts of AI. A focus on skills will be important both to help workers transition to a labour market where AI is widely used, and to support the diffusion of AI.
• Development of regulatory approaches for AI
  Globally, regulatory approaches to AI can broadly be divided between countries that have adopted comprehensive AI-specific legislation (in the European Union and China), and countries that rely on existing regulatory frameworks (the United Kingdom, the United States, Singapore, Japan). For New Zealand, existing regulatory frameworks, like copyright laws, may need to be updated to address the challenges of AI. Over time, aligning our regulations with other countries, where it makes sense for New Zealand, will be important to support the diffusion of AI.

This paper addresses AI more broadly, but also discusses generative AI specifically at points. The emergence of new forms of generative AI technology has been striking, perhaps most of all because it appears that many of the recent tools have been able to demonstrate rapid advances in higher-skill and creative tasks that were previously imagined to be harder to replicate, and so thought to be uniquely ‘human’. But the technologies underpinning this recent shift have been progressing for a number of years. And concern about the policy implications of new technologies is a classic topic in the history of economic thinking. This note therefore places AI in the context of that ongoing debate on technology.

As a general point of introduction, we note that the attitude of economists towards automation technology like AI has shifted in the past decade. In summarising this intellectual trajectory, leading labour economist David Autor writes: “the nuance of economic understanding has improved… yet, traditional economic optimism about the beneficent effect of technology for productivity and welfare has eroded as understanding has advanced.”[1] As New Zealand policymakers, this should lead us to take a more balanced view of the opportunities and challenges that might lie ahead as this technology continues to develop around the world.