Martin Fleming
Research Subjects
Research Interests
Martin Fleming has a total of 4 co-authors. Here are the top co-authors by number of shared publications:
Brian Goehring
2 shared publicationsMaja Svanberg
2 shared publications- Neil Thompson
Massachusetts Institute of Technology, Cambridge, USA
2 shared publications Wensu Li
2 shared publications
Latest Publications
- Open Access
- 1 May 2026
- Published by Elsevier in SSRN Electronic Journal
BEYOND AI EXPOSURE:WHICH TASKS ARE COST-EFFECTIVE TO AUTOMATE?
The faster artificial intelligence (AI) automation spreads through the economy, the more profound its potential impacts, both positive and negative. However, the existing literature on “AI Exposure” cannot predict the pace of automation since it attempts to measure an overall potential for AI, not the technical feasibility and economic attractiveness of building such systems. In this paper, a new type of AI task automation framework is presented. While the framework applies to generative AI models and deep learning predictive models, computer vision models illustrate the approach, where cost data are more well developed. We find that at today’s costs U.S. businesses would choose not to automate most vision tasks that have “AI Exposure,” and that only 23% of worker wages being paid for vision tasks would be attractive to automate. The slower roll-out of AI can be accelerated if costs falls rapidly or if it is deployed via AI agents or AI-as-a-service that have greater scale than individual firms. Overall, our findings suggest that AI job displacement will be substantial, but also gradual – and therefore there is room for policy and retraining to mitigate unemployment impacts.
- 1 May 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessBEYOND AI EXPOSURE:WHICH TASKS ARE COST-EFFECTIVE TO AUTOMATE?
The faster artificial intelligence (AI) automation spreads through the economy, the more profound its potential impacts, both positive and negative. However, the existing literature on “AI Exposure” cannot predict the pace of automation since it attempts to measure an overall potential for AI, not the technical feasibility and economic attractiveness of building such systems. In this paper, a new type of AI task automation framework is presented. While the framework applies to generative AI models and deep learning predictive models, computer vision models illustrate the approach, where cost data are more well developed. We find that at today’s costs U.S. businesses would choose not to automate most vision tasks that have “AI Exposure,” and that only 23% of worker wages being paid for vision tasks would be attractive to automate. The slower roll-out of AI can be accelerated if costs falls rapidly or if it is deployed via AI agents or AI-as-a-service that have greater scale than individual firms. Overall, our findings suggest that AI job displacement will be substantial, but also gradual – and therefore there is room for policy and retraining to mitigate unemployment impacts.
- Open Access
- 1 January 2024
- Published by Elsevier in SSRN Electronic Journal
Beyond AI Exposure: Which Tasks are Cost-Effective to Automate with Computer Vision?
The faster AI automation spreads through the economy, the more profound its potential impacts, both positive (improved productivity) and negative (worker displacement). The previous literature on “AI Exposure” cannot predict this pace of automation since it attempts to measure an overall potential for AI to affect an area, not the technical feasibility and economic attractiveness of building such systems. In this article, we present a new type of AI task automation model that is end-to-end, estimating: the level of technical performance needed to do a task, the characteristics of an AI system capable of that performance, and the economic choice of whether to build and deploy such a system. The result is a first estimate of which tasks are technically feasible and economically attractive to automate - and which are not. We focus on computer vision, where cost modeling is more developed. We find that at today’s costs U.S. businesses would choose not to automate most vision tasks that have “AI Exposure,” and that only 23% of worker wages being paid for vision tasks would be attractive to automate. This slower roll-out of AI can be accelerated if costs falls rapidly or if it is deployed via AI-as-a-service platforms that have greater scale than individual firms, both of which we quantify. Overall, our findings suggest that AI job displacement will be substantial, but also gradual – and therefore there is room for policy and retraining to mitigate unemployment impacts.
- 1 January 2024
- Published by Elsevier in SSRN Electronic Journal
Open AccessBeyond AI Exposure: Which Tasks are Cost-Effective to Automate with Computer Vision?
The faster AI automation spreads through the economy, the more profound its potential impacts, both positive (improved productivity) and negative (worker displacement). The previous literature on “AI Exposure” cannot predict this pace of automation since it attempts to measure an overall potential for AI to affect an area, not the technical feasibility and economic attractiveness of building such systems. In this article, we present a new type of AI task automation model that is end-to-end, estimating: the level of technical performance needed to do a task, the characteristics of an AI system capable of that performance, and the economic choice of whether to build and deploy such a system. The result is a first estimate of which tasks are technically feasible and economically attractive to automate - and which are not. We focus on computer vision, where cost modeling is more developed. We find that at today’s costs U.S. businesses would choose not to automate most vision tasks that have “AI Exposure,” and that only 23% of worker wages being paid for vision tasks would be attractive to automate. This slower roll-out of AI can be accelerated if costs falls rapidly or if it is deployed via AI-as-a-service platforms that have greater scale than individual firms, both of which we quantify. Overall, our findings suggest that AI job displacement will be substantial, but also gradual – and therefore there is room for policy and retraining to mitigate unemployment impacts.
- Open Access
- 18 May 2023
- Published by Elsevier in SSRN Electronic Journal
Enterprise Information and Communications Technology Software Pricing and Developer Productivity Measurement
The 1999 addition of business sector software and services spending was an important National Income and Product Accounts innovation, achieving a novel focus on the measurement of intangible asset investment. Over the intervening years, enterprise information and communication technology (ICT) has fundamentally changed. As a software producing sector, the business sector ICT function now has a much wider array of choices - software-as-a-service; open-source software; computing, storage, and communications cloud services; and developer services from a range of global resources. Labor and multifactor software development productivity are important sources of value creation, making measurement challenging. As the software development sector has become increasingly important in providing business sector ICT productivity, software sector developer productivity has become a viable proxy for business sector ICT developer productivity. With the use of a two-sector model and a standard growth accounting framework, the business sector ICT function shadow price is estimated, finding software price declines have been underestimated by 6.0 percentage points (ppt) over 2015 to 2020. The impact on GDP growth is a 0.1 ppt underestimate and the impact increases software spending from 19.1% to 24.9% of nonresidential fixed investment.
- 18 May 2023
- Published by Elsevier in SSRN Electronic Journal
Open AccessEnterprise Information and Communications Technology Software Pricing and Developer Productivity Measurement
The 1999 addition of business sector software and services spending was an important National Income and Product Accounts innovation, achieving a novel focus on the measurement of intangible asset investment. Over the intervening years, enterprise information and communication technology (ICT) has fundamentally changed. As a software producing sector, the business sector ICT function now has a much wider array of choices - software-as-a-service; open-source software; computing, storage, and communications cloud services; and developer services from a range of global resources. Labor and multifactor software development productivity are important sources of value creation, making measurement challenging. As the software development sector has become increasingly important in providing business sector ICT productivity, software sector developer productivity has become a viable proxy for business sector ICT developer productivity. With the use of a two-sector model and a standard growth accounting framework, the business sector ICT function shadow price is estimated, finding software price declines have been underestimated by 6.0 percentage points (ppt) over 2015 to 2020. The impact on GDP growth is a 0.1 ppt underestimate and the impact increases software spending from 19.1% to 24.9% of nonresidential fixed investment.