AI, Remote Work and Productivity: An International Conference

Chair: Benoit Dostie

Professeur, HEC Montréal, Directeur académique, CIQSS

Andrew Sharpe

Founder and Executive Director, Centre for the Study of Living Standards (CSLS)

“Can Canada Return to Trend Labour Productivity Growth of 1 Per Cent?”

Abstract:

Since 2019, labour productivity growth in Canada has been abysmal. Between 2019 and 2024 business sector output per hour advanced at only a 0.4 per cent average annual pace, compared to around 1 per cent from 2000 to 2019 and of course much higher productivity growth rates before 2000. The objective of this presentation is to examine the reasons for the productivity drought or emergency, as it has been called by the Bank of Canada. The presentation also assesses the chances of Canada retuning to the pre-2019 productivity growth trend, either with or without policies targeted to improve productivity. Particular attention will be paid to the potential of AI to boost productivity, and to the impact of remote work on productivity.

Arthur Sweetman

Professor of Economics, McMaster University

“Big Data and Long-term Care: Describing a New Platform in Ontario”

Abstract:

The impact of automation on productivity remains a subject of ongoing debate, with empirical studies yielding mixed results. Drawing on data from the 2019 and 2021 waves of the Korean Workplace Panel Survey (WPS), this study finds that automation has a positive effect on productivity. Our analysis highlights changes in occupational structure as a key mechanism linking automation to productivity gains. However, these gains are not evenly distributed: they vary significantly across industries and between unionized and non-unionized workplaces. Furthermore, the effect of automation is moderated by the intensity of worker training and the degree to which workers are involved in decision-making related to technology adoption.

Jocelyn Maclure

Stephen A. Jarislowsky Chair in Human Nature and Technology and Professor of Philosophy, McGill University

“Why is AI not Revolutionizing Work? The Epistemology and Ethics of Deep Artificial Neural Networks”

Abstract:

Deep artificial neural networks—including LLMs—demonstrate impressive capacities. They can extend human cognition in various ways, support human judgment and be turned into a suite of virtual assistants. Many inferred from these advances that machine learning-based applications would have a profound effect on the economy and the job market, increasing both productivity and unemployment. Yet, contra the prognostications of several economists and tech leaders, AI adoption is much slower than predicted, and its effects on employment are hard to discern. I will argue that this is mainly due to the fundamental epistemic limitations of the best deep learning systems, as well as to the ethical risks that their deployment raises. If this view is accurate, it suggests that AI is unlikely to lead to massive unemployment, as the orientation that should guide public and private organizations that seek to implement AI technologies should be to ensure genuine human control over AI.