Artificial Intelligence and Job Loss Peer Reviewed Journals
Proc Natl Acad Sci U South A. 2019 Apr 2; 116(fourteen): 6531–6539.
Economic Sciences
Toward agreement the bear on of artificial intelligence on labor
Morgan R. Frank
aMedia Laboratory, Massachusetts Institute of Technology, Cambridge, MA, 02139;
David Autor
bDepartment of Economic science, Massachusetts Found of Applied science, Cambridge, MA, 02139;
James Due east. Bessen
cTechnology & Policy Enquiry Initiative, School of Law, Boston University, Boston, MA, 02215;
Erik Brynjolfsson
dSloan School of Management, Massachusetts Constitute of Engineering science, Cambridge, MA, 02139;
eNational Bureau of Economical Research, Cambridge, MA, 02138;
Manuel Cebrian
aMedia Laboratory, Massachusetts Constitute of Technology, Cambridge, MA, 02139;
David J. Deming
fHarvard Kennedy School, Harvard Academy, Cambridge, MA, 02138;
gGraduate School of Education, Harvard University, Cambridge, MA, 02138;
Maryann Feldman
hDepartment of Public Policy, The Academy of North Carolina at Chapel Hill, Chapel Hill, NC, 27599;
Matthew Groh
aMedia Laboratory, Massachusetts Institute of Technology, Cambridge, MA, 02139;
José Lobo
iSchool of Sustainability, Arizona Land University, Tempe, AZ, 85287;
Esteban Moro
aMedia Laboratory, Massachusetts Institute of Technology, Cambridge, MA, 02139;
jGrupo Interdisciplinar de Sistemas Complejos, Departmento de Matematicas, Escuela Politécnica Superior, Universidad Carlos 3 de Madrid, 28911 Madrid, Kingdom of spain;
Dashun Wang
grandKellogg School of Management, Northwestern Academy, Evanston, IL, 60208;
lNorthwestern Found on Complex Systems, Northwestern University, Evanston, IL, 60208;
Hyejin Youn
kKellogg School of Management, Northwestern University, Evanston, IL, 60208;
lNorthwestern Constitute on Complex Systems, Northwestern University, Evanston, IL, 60208;
Iyad Rahwan
aMedia Laboratory, Massachusetts Institute of Applied science, Cambridge, MA, 02139;
mPlant for Data, Systems, and Gild, Massachusetts Constitute of Technology, Cambridge, MA, 02139;
nEye for Humans and Machines, Max Planck Institute for Human Development, 14195 Berlin, Frg
Abstract
Rapid advances in artificial intelligence (AI) and automation technologies have the potential to significantly disrupt labor markets. While AI and automation can augment the productivity of some workers, they tin replace the work done by others and will likely transform nigh all occupations at least to some degree. Rise automation is happening in a period of growing economical inequality, raising fears of mass technological unemployment and a renewed phone call for policy efforts to address the consequences of technological change. In this paper nosotros hash out the barriers that inhibit scientists from measuring the effects of AI and automation on the hereafter of work. These barriers include the lack of high-quality data about the nature of work (eastward.g., the dynamic requirements of occupations), lack of empirically informed models of cardinal microlevel processes (e.1000., skill substitution and human–automobile complementarity), and insufficient understanding of how cerebral technologies interact with broader economic dynamics and institutional mechanisms (e.g., urban migration and international trade policy). Overcoming these barriers requires improvements in the longitudinal and spatial resolution of data, likewise every bit refinements to data on workplace skills. These improvements will enable multidisciplinary enquiry to quantitatively monitor and predict the complex evolution of work in tandem with technological progress. Finally, given the fundamental doubtfulness in predicting technological modify, we recommend developing a determination framework that focuses on resilience to unexpected scenarios in addition to general equilibrium behavior.
Keywords: automation, employment, economic resilience, time to come of piece of work
Bogus Intelligence (AI) is a speedily advancing form of technology with the potential to drastically reshape US employment (1, 2). Unlike previous technologies, examples of AI accept applications in a variety of highly educated, well-paid, and predominantly urban industries (3), including medicine (4, 5), finance (6), and information technology (7). With AI's potential to alter the nature of piece of work, how can policy makers facilitate the next generation of employment opportunities? Studying this question is made hard by the complication of economical systems and AI'southward differential impact on unlike types of labor.
While engineering generally increases productivity, AI may diminish some of today'south valuable employment opportunities. Consequently, researchers and policy makers worry well-nigh the time to come of work in both advanced and developing economies worldwide. As an example, China is making AI-driven applied science the centerpiece of its economical development program (8). Automation concerns are not new to AI, and examples date back even to the advent of written language. In ancient Greece (ca. 370 BC), Plato's Phaedrus (9) described how writing would displace human retentiveness and reading would substitute true knowledge with mere data. More ordinarily, historians bespeak to the Industrial Revolution and the riots of xixth-century Luddites (x) equally examples where technological advancement led to social unrest. Ii examples from the recent past echo these concerns.
Wassily Leontief, winner of the 1973 Nobel Prize in Economics, noted in 1952, "Labor will become less and less important. . . More than workers volition be replaced by machines. I exercise not see that new industries can employ everybody who wants a job" (xi).
Similarly, US Attorney General Robert F. Kennedy commented in 1964, "Automation provides u.s. with wondrous increases of production and information, only does it tell us what to do with the men the machines readapt? Modern manufacture gives u.s.a. the capacity for unparalleled wealth—merely where is our capacity to brand that wealth meaningful to the poor of every nation?" (12).
However, despite these long-lasting and oft-recurring concerns, society underwent profound transformations, the economy continued to grow, engineering science continues to advance, and workers continue to have jobs. Given this history of business, what makes human labor resilient to automation? Is AI a fundamentally new business organisation from technologies of the past?
Answering these questions requires a detailed cognition that connects AI to today'southward workplace skills. Each specific engineering science alters the need for specific types of labor, and thus the varying skill requirements of different job titles can obfuscate applied science's bear upon. In full general, depending on the nature of the job, a worker may be augmented by technology or in contest with information technology (13–15). For example, technological advancements in robotics can diminish wages and employment opportunities for manufacturing workers (xvi, 17). However, technological change does non necessarily produce unemployment, and, in the case of AI, cognitive engineering may actually augment workers (18, 19). For instance, machine learning appears to bolster the productivity of software developers while also creating new investment and manufacturing opportunities (e.g., autonomous vehicles). Complicating matters further, the skill requirements of occupations exercise non remain static, but instead change with changing engineering (nineteen, 20).
In the remainder of this article, we describe how these competing dynamics combined with insufficient data might allow contrasting perspectives to coexist. In item, we fence that the limitations into data about workplace tasks and skills restricts the viable approaches to the trouble of technological alter and the futurity of work. We offer suggestions to improve information collection with the goal of enriching models for workplace skills, employment, and the impact of AI. Finally, we suggest insights that improved information could provide in combination with a methodological focus on resilience and forecasting.
Contrasting Perspectives
Doomsayer's Perspective.
Engineering improves to make man labor more than efficient, just large improvements may yield deleterious effects for employment. This obsoletion through labor commutation leads many to worry near "technological unemployment" and motivates efforts to forecast AI's bear on of jobs. One study assessed recent developments in AI to conclude that 47% of electric current U.s. employment is at high risk of computerization (23), while a contrasting study, using a different methodology, ended that a less alarming 9% of employment is at hazard (24). Similar studies have looked at the impact of automation on employment in other countries and reached sobering conclusions: Automation will affect 35% of employment in Republic of finland (25), 59% of employment in Germany (26), and 45 to 60% of employment across Europe (27). Critics have complained that prospective studies lack validation, but retrospective studies also find that robotics are diminishing employment opportunities in US manufacturing (17, 28) [although not in Germany (29)].
Optimist's Perspective.
Optimists suggest that technology may substitute for some types of labor but that efficiency gains from technological augmentation outweigh transition costs (30–34), and, in many cases, technology increases employment for workers who are in non direct competition with information technology (19, 35) [although recent follow-up work suggests these are temporary gains (28)]. Furthermore, the skill requirements of each job title are non static and actually evolve over time to reflect evolving labor needs. For case, workers may require more social skills because those skills remain difficult to automate (xx). Even if technology depresses employment for some types of labor, it tin create new needs and new opportunities through "creative devastation" (36–38). For example, the replacement of equestrian travel with automobiles spurred demand for new roadside amenities, such as motels, gas stations, and fast food (39).
Unifying Perspectives.
On one mitt, multiple dynamics back-trail technological change and create doubt about the future of work. On the other paw, experts agree that occupations are best understood equally abstract bundles of skills (18, twoscore) and that engineering directly impacts demand for specific skills instead of acting on whole occupations all at in one case (16, 19, 35, 41). Therefore, a detailed framework that connects specific skill types to career mobility (18, 42) and to whole urban workforces (40) may help to unify competing perspectives (Fig. aneC ). Existing studies have argued theoretically that unlike skill types underpin aggregate labor trends, such equally job polarization (16) and urban migration (43, 44), simply robust empirical validation is made difficult past the specificity of mod skills information and their temporal sparsity.
Motivating and describing a framework to written report technology's impact on workplace skills. (A) Following ref. 21, we apply American Community Survey national employment statistics to compare the modify in employment share (y centrality) of occupations according to their average almanac wage (ten axis) during two time periods. Employment share is increasing for low- and high-wage occupations at the expense of middle-wage occupations. (B) Following ref. 15, we use data from the Federal Reserve Bank of St. Louis to compare US productivity (real output per hour) and workers' income (real median personal income), which have traditionally grown in tandem. The efficiency gains of automating technologies are thought to contribute to this and so-called neat decoupling starting around the year 2000. (C) A framework for studying technological change, workplace skills, and the futurity of work every bit multilayered network. (Left) Cities and rural areas represent separate labor markets, but workers and appurtenances can flow betwixt them. (Center) Each location tin can be represented as an employment distribution across occupations. Connections between occupations in a labor market represent viable job transitions. Job transitions are viable if workers of one job can encounter the skill requirements of another job [i.east., "skill matching" (22)]. (Right) Workers' varying skill sets represent bundles of workplace skills that tend to exist valuable together. Skill pairs that tend to cooccur may identify paths to career mobility. Applied science alters demand for specific workplace skills, thus altering the connections betwixt skill pairs. Every bit an example, machine vision software may impact the demand for homo labor for some visual task. These alterations can accrue and diffuse throughout the entire arrangement every bit aggregate labor trends described in A and B.
Overcoming Barriers to Forecasting the Future of Work
In this department nosotros identify barriers to our scientific modeling of technological modify and the future of work. Along with each barrier, nosotros propose a potential solution that could enable improvement in forecasting labor trends. We provide a summary of these barriers and solutions in Table 1.
Table 1.
Tabulating the current barriers to forecasting the future of work forth with proposed solutions
| Bulwark | Potential solution |
| Sparse skills data |
|
| Limited modeling of resilience |
|
| Places in isolation |
|
Bulwark: Sparse Skills Data.
Forecasting automation from AI requires skills information that keep stride with chop-chop advancing engineering [e.g., Moore'south Law (45), robots in manufacturing (17), and patent production (46–48)]. While skill types inform the theory of labor and technological change (1, 18, 21, 49), standard labor data focus on aggregate statistics, such as wage and employment numbers, and tin can lack resolution into the specifics that distinguish different chore titles and dissimilar types of work. For example, previous studies have empirically observed a "hollowing" of the middle-skill jobs described by increasing employment share for low-skill and high-skill occupations at the expense of middle-skill occupations (16, 35) (reproduced in Fig. oneA ). These studies use skills to explain labor trends only are limited empirically to measuring almanac wages instead of skill content straight. While wages may correlate with specific skills, wage alone fails to capture the defining features of an occupation, and models focused on only cognitive and physical labor fail to explain responses to technological alter (21).
As another approach, data on educational requirements can add resolution to employment trends (fifty–52). For instance, jobs that crave a bachelor's degree may identify cognitive workers who are less susceptible to automation. Ideally, educational institutions railroad train workers to possess valuable skills that lead to higher wages (53). Yet, looking at didactics and wages alone has proven insufficient to explain stagnating returns on didactics (xvi, 54, 55) and slow wage growth despite increases in national productivity (14, 15, 41) (Fig. aneB ).
Improving data on the skills required to perform specific job tasks may provide improve insights than wages and education solitary. For example, previous studies have considered occupations as routine or nonroutine and cognitive or physical (21, 56–63) or looked at specific types of skills in relation to augmentation and substitution from technology (18, 41). Increasing a labor model'south specificity into workplace tasks and skills might further resolve labor trends and improve predictions of automation from AI. Every bit an example, consider that ceremonious engineers and medical doctors are both loftier-wage, cognitive, nonroutine occupations requiring many years of higher education and additional professional certification. Yet, these occupations crave distinct workplace skills that are largely nontransferable, and these occupations are likely to collaborate with different technologies. Wages and pedagogy—and even aggregations of workplace skills—may be besides coarse to distinguish occupations and, thus, may obfuscate the differential impact of diverse technologies and complicate predictions of changing skill requirements. In turn, these shortcomings may help explain the variability in current automation predictions that enable contrasting perspectives.
While publicly available skills data are limited, the U.s. Section of Labor's O*NET database has seen contempo apply in labor inquiry (e.g., refs. 23, 41, and 64). O*NET offers many benefits including a detailed taxonomy of skills and more than regular updates than preceding datasets. In 2014, O*NET began to receive partial updates twice annually, which is a considerable comeback on the Dictionary of Occupational Titles, which was published in four editions in 1939, 1949, 1965, and 1977, with a revision in 1991. Still, employment trends and changing demand for specific tasks and skills might change faster than O*Internet'southward temporal resolution and skill categorization tin can capture. Complicating matters farther, advances in AI and machine learning may be changing the nature of automation, thereby altering the types of tasks that are affected by technology (3, 65).
Furthermore, studies often use O*NET information to construct aggregations of skills, such equally information input or mental processes (40), rather than focusing on skills at their near granular level. Methodological choices aside, O*NET'due south relatively static skill taxonomy poses its own issues as well. For instance, according to O*Internet, the skill "installation" is equally of import to both estimator programmers and to plumbers, but, undoubtedly, workers in these occupations are performing very unlike tasks when they are installing things on the job (see Fig. 2A and SI Appendix, section 1 for calculation). More generally, any static taxonomy for workplace skills is non ideal for a changing economy: Should mathematics and programming exist two separate workplace skills given that they are both computational? Conversely, is "programming" likewise wide given the variety of existing software and programming languages? Perchance it is more appropriate to specify programming tasks or specific programming languages (see Fig. 2B for an example), specially given the rapid development of AI and machine learning. Likely, the correct abstraction is situation-dependent, but O*NET data offering limited flexibility.
Since the skill requirements of occupations may inform opportunities for career mobility, abstract skill data may obfuscate important labor trends. (A) We use O*Cyberspace information to place the characteristic skill requirements for truck drivers, plumbers, and software developers (see SI Appendix, section one for calculation). Private skills may be unique to an occupation (eastward.chiliad., operating vehicles) or shared between occupations (e.thousand., depression-light vision). The skill of installation is required past both plumbers and software developers, but this skill may non mean the same thing to workers in these 2 occupations. Programming is a skill required by software developers, simply the coarseness of this skill definition may hide of import dynamics brought on past new technology, including AI. (B) For example, we provide the percent of Google searches for coding tutorials by programming linguistic communication. Trends are smoothed using locally weighted scatter plot smoothing (run into SI Appendix, section two for calculation). The Python programming language is widespread in the field of motorcar learning. Therefore, the increased ubiquity of AI and, in detail, car learning may contribute to Python'due south steady growth in popularity.
Granular skills information will help elucidate the micro-scale touch on of AI and other technologies in labor systems. For instance, the specifications of contempo patents might propose automatable types of labor in the near future (46–48), thus elucidating the impact of technological change at the granularity of workplace-specific tasks and skills. The distribution of skill categories inside occupations and over individuals' careers can reveal how occupational skill requirements evolve. Every bit an case, consider that occupations such as software programmer dynamically change the skill requirements in job listings (e.m., "programming" in the 1990s vs. "Python," "Java," "Kubernetes," etc. today) to reflect the tools and required specialization of the fourth dimension. Agreement the dynamics of specific skills combined with the incomes within occupations can capture the marginal value of different skills despite the dynamic nature of piece of work.
Online career platforms offer an instance of the empirical possibilities facilitated by nontraditional and new data sources. These websites collect real-time information that reflect labor dynamics in certain industries. Information from workers' resumes can improve our agreement of education and careers, likewise as identifying workers' transitions between occupations and skill sets. Additionally, chore postings capture fluctuations in labor demands and demonstrate changes in demand for specific skills. Combined, these two sources of skills information offering an adaptive granular view into the changing nature of work that may item where labor disconnects exist. Access to these private information sources is currently restricted and typically requires a data-sharing agreement that protects personally identifiable information and other proprietary information. Of grade, personal privacy and issues of representative sampling are inherent to these data, but increased access could meaningfully augment currently available open up data on employment and workplace skills. I potential solution is to construct a secure environment for the sharing of detailed skills and career information that is similar to the contempo Social Science 1 partnership (69) (see https://socialscience.one).
Barrier: Express Modeling of Resilience.
Contempo studies show that historical technology-driven trends may not capture the AI-driven trends we face today. Consequently, some have concluded that AI is a fundamentally new technology (3, 65). If the trends of the past are non predictive of the employment trends from current or time to come technologies, and then how can policy makers maintain and create new employment opportunities in the face of AI? What features of a labor marketplace lead to generalized labor resilience to technological change?
Information technology is difficult to construct resilient labor markets considering of the uncertainty effectually engineering science's impact on labor. For instance, designing viable worker retraining programs requires detailed cognition of the local workforce, fluency with current technology, and an agreement of the complex dependencies between regional labor markets around the world (70, 71). Engineering science typically performs specific tasks and may change demand for specific workplace skills as a result. These micro-scale changes to skill demand can accrue into systemic labor trends including occupational skill redefinition, employment redistribution (e.chiliad., task creation and technological unemployment), and geographic redistribution (eastward.g., worker migration). Forecasting these complex effects requires a detailed understanding of the pathways forth which these dynamics occur.
Equally an emblematic example of these complex dynamics, consider the competition between human bank tellers and automated teller machines (ATMs) (described in ref. 72). Unexpectedly, national employment for banking concern tellers rose with the adoption of ATMs. One caption is demand elasticity: As ATMs decreased the operating cost of bank branches, more bank branches opened nationwide to see rising consumer demand. Another more complicated reason is the accompanying shift in cardinal skill requirements from clerical ability to social and persuasive skills used by salespeople and customer service representatives. The story of bank tellers and ATMs is simply fully captured by connecting the job-level changes in occupational skill limerick with the arrangement-level dynamics of need brought on by increased efficiency. Accordingly, an updated framework for labor and AI must capture the interactions of microscopic workplace skills in combination to produce macroscopic labor trends, such as employment shifts, job polarization, and workers' spatial mobility (for instance, see Fig. 3B ).
Skill complementarity may define the structural resilience of a workforce and inform worker retraining programs. (A) As in climatology and ecology, the structural pathways constraining labor dynamics could determine the resilience of a labor market place to changing labor skill demands. In this instance, we connect occupation pairs with high skill similarity considering skill similarity might point easier worker transitions between job titles. Borrowing from research on ecological systems (66), the density of connections betwixt occupations could determine "tipping points" for aggregate employment in cities. (B) With contempo concerns of automation (67, 68), which jobs might be suitable for paralegals and legal assistants if employment for these jobs diminishes? Ameliorate resolution into skill requirements could aid identify occupations that rely on similar skills but too rely on skills that are removed from competition with technology. In this example, nosotros identify characteristic skills using the O*NET database to find that paralegals rely on many shared workplace skills with homo resource specialists. Human resource specialists rely on social skills, which are not easily automatic (twenty). See SI Appendix, section 1 for skill calculations.
Existing theory of the matching process between task seekers and job vacancies (22) provides a stylized description of the matching process that lacks resolution into skill need. Mapping the infinite of skill interdependencies (east.one thousand., Fig. 1C ) could inform grooming and job assistance programs past identifying which types of work—and which locations—may experience augmentation and/or commutation with new technology. The detailed skill requirements of occupations decide the career mobility of individual workers, and thus changes to the need for sure skills have the potential to redefine feasible career trajectories and worker flow between occupations (e.g., heart layer of Fig. 1D ). Therefore, mapping the relationships between jobs and skills that produce employment opportunities is a vital pace for policy makers in the face of technological modify.
In related domains, tools from network scientific discipline have already provided new insights into modeling (and minimizing) systemic risk (73) in global credit (74) and financial industries (75), forecasting the time to come exports of national economies (76–78), mapping worker flows between industries (79) and firms (80), and charting the changing industrial composition of cities (81–83) and municipalities (84). Therefore, identifying the pathways forth which labor dynamics (e.g., how skills determine workers' career mobility) occur may provide similarly useful insights into the impact of AI on labor. Similar methods have been used to measure ecological resilience based on the structure of mutualistic interspecies interactions (66, 85). These methods often rely on the size and density of interconnected entities to estimate systemic resilience to species removal—peradventure analogous to diminishing need for a skill with new engineering (due east.m., Fig. 3A ).
Mapping skill dependencies volition require advisable data-treatment methods. The ideal skills data should reflect the dynamic nature of skill representation, then the methods we use to find, categorize, and mensurate the demand for skills must be adjustable as well. Peradventure ironically, advanced AI techniques may assist. Tools from machine learning (ML) and natural language processing (NLP) may capture the latent structure in complex high-dimensional data, thus making them platonic tools for the proposed application [and other applications in econometrics (86)]. For example, NLP may be used to process historical skills information from the Dictionary of Occupational Titles into a format alike to the modern O*Cyberspace data. ML can exist used on longitudinal chore postings data to identify trends in skill demands that may reflect changes in technological ability. Combining these modern computational methods with relevant sources of data may foster new insights into labor dynamics at a loftier temporal resolution. In plow, these methodological improvements tin can bolster labor forecasts and policy makers' power to respond to real-time labor trends.
Barrier: Places in Isolation.
The touch of AI and automation will vary greatly beyond geography, which has implications for the labor forcefulness, urban–rural discrepancies, and changes in the income distribution (87). The study of AI and automation are largely focused on national employment trends and national wealth disparity. However, recent work demonstrates that some places (e.grand., cities) are more susceptible to technological change than others (17, 64). Occupations form a network of dependencies which constrain how easily jobs tin can exist replaced by applied science (82, 88). Therefore, the wellness of the amass labor market place may depend on the affect of applied science on specific urban and rural labor markets (73, 84).
Although technological modify alters demand for specific workplace tasks and skills, current skills data mask the specific skill sets that contain and differentiate the workforces of dissimilar geographies. In part, this is because skills data from nationwide surveys, such as the O*NET database, boilerplate over the regional variability in the required skills of workers with shared job titles. For example, software developers seeking employment in Silicon Valley may need to advertise more than specific skill sets than similar employees in a shallower labor market place (post-obit the division of labor theory). Exacerbating this trend, the same AI technologies that broaden high-wage cognitive employment are more abundant in large cities, while the physical low-wage tasks that are most readily replaced by robotics are more abundant in pocket-sized cities and rural communities. This observation suggests that national wealth disparity is reflected in the wealth disparity between large and small cities akin to wage inequality across individuals.
Improved models for spatial interdependencies require more granular skills data (discussed above) and new insights into the mechanisms that create today'south cross-sectional geographic trends. For instance, how exercise university towns, where people proceeds valuable cognitive skills, contribute to the productivity of large cities? Practice these economic connections help explain why university towns perform surprisingly well compared with similarly sized cities according to socioeconomic indicators [including exposure to automation (64)]?
Furthermore, but as internal connectivity determines urban economical resilience (83), so too can the connections betwixt US cities underpin the economic wellness of the national economic system (48). For instance, an intermission in the supply chain of well-educated cognitive workers may stifle an urban economy that normally attracts skilled workers. Therefore, it behooves policy makers to empathize the connections between their local labor market and other urban labor markets to appraise the resilience of their local economy. Since employment opportunities are central in people'south decision to relocate (43) and skill matching is essential to the job matching procedure (22), agreement the constituent skill sets in cities tin inform models for the spatial mobility of workers and improve our agreement of career mobility and career incentives.
Determination
AI has the potential to reshape skill demands, career opportunities, and the distribution of workers among industries and occupations in the United States and in other adult and developing countries. However, researchers and policy makers are underequipped to forecast the labor trends resulting from specific cerebral technologies, such as AI. Typically, technology is designed to perform a specific task which alters demand for specific workplace skills. The resulting alterations to skill demands diffuse throughout the economy, influencing occupational skill requirements, career mobility, and societal well-being (e.thousand., impacts to workers' social identity). Identifying the specific pathways of these dynamics has been constrained by coarse historical data and limited tools for modeling resilience. We can overcome these obstacles, however, by prioritizing data collection that is detailed, responsive to existent-time changes in the labor market, and respects regional variability (see Fig. four for a information-pipeline schematic). Specifically, better access to unstructured skills data from resumes and job postings along with new indicators for recent technological change (due east.chiliad., patent information) and models for both intercity and intracity labor dependencies will enable new and promising techniques for understanding and forecasting the futurity of work. This improved information collection will enable the apply of new data-driven tools, including machine learning applications and systemic modeling that more accurately reflects the complexity of labor systems. New data will pb to new inquiry that enriches our agreement of the affect of technology on modern labor markets.
A data pipeline that overcomes barriers to studying the future of piece of work. (A) Inputs into the data pipeline include structured and unstructured data that detail regional variations in labor and granular skills data in relation to technological alter. (B) Information from a variety of sources will demand to exist centralized and processed into a class that economists and information scientists can easily use (east.g., NLP to identify skill from resume and job postings). (C) Cleaned data feed a model for both the intercity (e.m., worker migration) and intracity (eastward.g., changes to local career mobility) labor trends brought on past technological change. (D) Outputs from this model will forecast the labor impact of technological change. These forecasts will inform policy makers seeking to implement prudent policy and individual workers attempting to navigate their careers.
Supplementary Fabric
Supplementary File
Acknowledgments
This work summarizes insights from the workshop on Innovation, Cities, and the Future of Work, which was funded by NSF Grant 1733545. This work was supported by the Massachusetts Institute of Technology (MIT) and the MIT Initiative on the Digital Economic system.
Footnotes
References
1. Mitchell T, Brynjolfsson E. Track how engineering science is transforming piece of work. Nature. 2017;544:290–292. [PubMed] [Google Scholar]
two. Brynjolfsson E, Stone D, Syverson C. 2017. Artificial intelligence and the modern productivity paradox: A clash of expectations and statistics. NBER Working Paper 22401 (National Bureau of Economic Research, Cambridge, MA)
iii. Brynjolfsson E, Mitchell T. What tin can machine learning do? Workforce implications. Science. 2017;358:1530–1534. [PubMed] [Google Scholar]
4. Barlow 1000. AI and Medicine. O'Reilly Media; Sebastopol, CA: 2016. [Google Scholar]
5. Darcy AM, Louie AK, Roberts LW. Motorcar learning and the profession of medicine. JAMA. 2016;315:551–552. [PubMed] [Google Scholar]
half dozen. Dunis CL, Middleton Pw, Karathanasopolous A, Theofilatos Grand. Artificial Intelligence in Financial Markets: Cut Border Applications for Risk Management, Portfolio Optimization and Economics. Springer; New York: 2016. [Google Scholar]
7. Witten IH, Frank Eastward, Hall MA, Pal CJ. Information Mining: Practical Machine Learning Tools and Techniques. Morgan Kaufmann; Burlington, MA: 2016. [Google Scholar]
8. He ABY. 2017. How China is preparing for an AI-powered time to come (Wilson Center, Washington, DC)
nine. Yunis H. Plato: Phaedrus. Cambridge Univ Press; Cambridge, United kingdom of great britain and northern ireland: 2011. [Google Scholar]
x. Jones SE. Confronting Applied science: From the Luddites to Neo-Luddism. Routledge; Abingdon, UK: 2013. [Google Scholar]
11. Leontief W. Machines and man. Sci Am. 1952;187:150–164. [Google Scholar]
13. Acemoglu D, Autor D. Handbook of Labor Economic science. Vol four. Elsevier; Amsterdam: 2011. Skills, tasks and technologies: Implications for employment and earnings; pp. 1043–1171. [Google Scholar]
14. Brynjolfsson E, McAfee A. Race Against the Car: How the Digital Revolution Is Accelerating Innovation, Driving Productivity, and Irreversibly Transforming Employment and the Economy. Digital Frontier; Lexington, MA: 2012. [Google Scholar]
15. Brynjolfsson E, McAfee A. The Second Car Age: Work, Progress, and Prosperity in a Time of Bright Technologies. Norton; New York: 2014. [Google Scholar]
16. Autor D. 2010. The polarization of chore opportunities in the US labor marketplace: Implications for employment and earnings (Center for American Progress and The Hamilton Projection)
17. Acemoglu D, Restrepo P. 2017. Robots and jobs: Evidence from US labor markets. NBER Working Newspaper 23285 (National Agency of Economic Inquiry, Cambridge, MA)
18. Autor DH, Levy F, Murnane RJ. The skill content of recent technological change: An empirical exploration. Q J Econ. 2003;118:1279–1333. [Google Scholar]
19. Bessen JE. 2016. How calculator automation affects occupations: Technology, jobs, and skills. Law and Economics Enquiry Paper 15-49 (Boston Univ Schoolhouse of Law, Boston)
20. Deming D, Kahn LB. Skill requirements beyond firms and labor markets: Testify from job postings for professionals. J Labor Econ. 2018;36:S337–S369. [Google Scholar]
21. Acemoglu D, Autor D. Handbook of Labor Economic science. Vol four. Elsevier; Amsterdam: 2011. Skills, tasks and technologies: Implications for employment and earnings; pp. 1043–1171. [Google Scholar]
22. Mortensen DT, Pissarides CA. Task creation and job devastation in the theory of unemployment. Rev Econ Stud. 1994;61:397–415. [Google Scholar]
23. Frey CB, Osborne MA. The future of employment: How susceptible are jobs to computerisation? Technol Forecast Soc Change. 2017;114:254–280. [Google Scholar]
24. Arntz M, Gregory T, Zierahn U. 2016 The risk of automation for jobs in OECD countries: A comparative analysis. OECD Social, Employment and Migration Working Paper 189 (OECD Publishing, Paris). Bachelor at www.ifuturo.org/sites/default/files/docs/automation.pdf.
26. Brzeski C, Burk C. Die Roboter kommen: Folgen der Automatisierung für den deutschen Arbeitsmarkt [The Robots are coming: Consequences of Automation for the German Labor Market] INGDiBa Econ Res. 2015;30:ane–vii. German. [Google Scholar]
28. Bessen J. 2018 AI and Jobs: The role of need. NBER Working Paper No. 24235. Available at https://www.nber.org/papers/w24235. Accessed March 15, 2019.
29. Dauth Due west, Findeisen Southward, Südekum J, Woessner N. 2017. German robots – The impact of industrial robots on workers. IAB Discussion Paper 30/2017 (Constitute for Employment Research, Nuremberg, Germany)
31. Frase P. Four Futures: Life after Capitalism. Verso Books; New York: 2016. [Google Scholar]
35. David H, Dorn D. The growth of depression-skill service jobs and the polarization of the US labor market. Am Econ Rev. 2013;103:1553–1597. [Google Scholar]
36. Aghion P, Howitt P. 1990. A model of growth through creative destruction. NBER Working Paper 3223 (National Bureau of Economic Research, Cambridge, MA)
37. Aghion P, Howitt P. Growth and unemployment. Rev Econ Stud. 1994;61:477–494. [Google Scholar]
38. Bartelsman Due east, Haltiwanger J, Scarpetta Southward. 2004. Microeconomic testify of artistic destruction in industrial and developing countries (World Bank, Washington, DC)
40. Kok S, ter Weel B. Cities, tasks, and skills. J Reg Sci. 2014;54:856–892. [Google Scholar]
41. MacCrory F, Westerman G, Alhammadi Y, Brynjolfsson E. 2014 Racing with and against the car: Changes in occupational skill limerick in an era of rapid technological advance. Proceedings of the Xxx-Fifth International Conference on Information Systems. Bachelor at https://dblp.org/rec/html/conf/icis/MacCroryWAB14. Accessed March 15, 2019.
42. Alabdulkareem A, Frank MR, Lord's day L, AlShebli B, Hidalgo C, Rahwan I. Unpacking the polarization of workplace skills. Sci Adv. 2018;4:eaao6030. [PMC free commodity] [PubMed] [Google Scholar]
43. International Organisation for Migration 2015. Globe migration study 2015 - Migrants and cities: New partnerships to manage mobility (International System for Migration, Le G-Saconnex, Switzerland)
44. Fields GS. Rural-urban migration, urban unemployment and underemployment, and task-search activity in LDCs. J Dev Econ. 1975;2:165–187. [PubMed] [Google Scholar]
46. Youn H, Strumsky D, Bettencourt LMA, Lobo J. Invention as a combinatorial process: Show from The states patents. J R Soc Interf. 2015;12:20150272. [PMC costless article] [PubMed] [Google Scholar]
47. Strumsky D, Lobo J. Identifying the sources of technological novelty in the process of invention. Res Pol. 2015;44:1445–1461. [Google Scholar]
48. Strumsky D, Thill J-C. Profiling U.s. metropolitan regions by their social research networks and regional economic performance. J Reg Sci. 2013;53:813–833. [Google Scholar]
49. Weaver A, Osterman P. Skill demands and mismatch in U.s. manufacturing. ILR Rev. 2017;seventy:275–307. [Google Scholar]
fifty. Autor DH. Why are there still so many jobs? The history and future of workplace automation. J Econ Perspect. 2015;29:iii–xxx. [Google Scholar]
51. Goldin L, Katz C. The Race Between Education and Technology. Harvard Univ Press; Cambridge, MA: 2008. [Google Scholar]
52. Autor DH, Katz LF, Kearney MS. Trends in US wage inequality: Revising the revisionists. Rev Econ Stat. 2008;xc:300–323. [Google Scholar]
53. Kerckhoff AC. Education and social stratification processes in comparative perspective. Sociology of Education. 2001;74:3–18. [Google Scholar]
55. Beaudry P, Green DA, Sand BM. The great reversal in the need for skill and cognitive tasks. J Labor Econ. 2016;34:S199–S247. [Google Scholar]
56. Spitz-Oener A. Technical change, job tasks, and rising educational demands: Looking exterior the wage structure. J Labor Econ. 2006;24:235–270. [Google Scholar]
57. Antonczyk D, Fitzenberger B, Leuschner U. 2009 Can a task-based arroyo explicate the recent changes in the German wage structure? ZEW Word Papers 08-132 (ZEW – Leibniz Centre for European Economic Research, Mannheim, Federal republic of germany). Available at hdl.handle.net/10419/27615.
58. Dustmann C, Ludsteck J, Schanberg U. Revisiting the German wage construction. Q J Econ. 2009;124:843–881. [Google Scholar]
59. Firpo S, Fortin NM, Lemieux T. 2011 Occupational tasks and changes in the wage construction. IZA Discussion Papers 5542 (Found for the Study of Labor, Bonn, Deutschland). Bachelor at ftp.iza.org/dp5542.pdf.
60. Ikenaga T. Polarization of the japanese labor market: The adoption of it and changes in tasks contents. The Japanese Periodical of Labour Studies. 2009;584:73–90. [Google Scholar]
61. Michaels G, Natraj A, Van Reenen J. 2010. Has ICT polarized skill demand? Evidence from xi countries over 25 years. NBER Working Paper 16138 (National Bureau of Economical Research, Cambridge, MA)
62. Black SE, Spitz-Oener A. 2010. Explaining women'southward success: Technological change and the skill content of women'south work. NBER Working Newspaper 13116 (National Bureau of Economic Inquiry, Cambridge, MA)
63. Ikenaga T, Kambayashi R. 2010 Long-term trends in the polarization of the Japanese labor market: The increment of not-routine task input and its valuation in the labor market. Available at http://cis.ier.striking-u.ac.jp/Mutual/pdf/dp/2009/dp464_2.pdf. Accessed March fifteen, 2019.
64. Frank MR, Sun L, Cebrian M, Youn H, Rahwan I. Small cities face greater impact from automation. J R Soc Interf. 2018;15:20170946. [PMC free article] [PubMed] [Google Scholar]
65. Brynjolfsson E, Mitchell T, Rock D. AEA Papers and Proceedings. Vol 108. American Economic Association; Nashville, TN: 2018. What can machines learn, and what does it mean for occupations and the economy? pp. 43–47. [Google Scholar]
66. Gao J, Barzel B, Barabási A-Fifty. Universal resilience patterns in circuitous networks. Nature. 2016;530:307–312. [PubMed] [Google Scholar]
67. Simpson B. Algorithms or advocacy: Does the legal profession have a hereafter in a digital globe? Inf Commun Technol L. 2016;25:50–61. [Google Scholar]
68. Davenport Th, Kirby J. But Humans Demand Apply: Winners and Losers in the Age of Smart Machines. Harper Business; New York: 2016. [Google Scholar]
lxx. Walter Due south. Globalization and the demand-side of politics: How globalization shapes labor marketplace gamble perceptions and policy preferences. Polit Sci Res Methods. 2017;5:55–80. [Google Scholar]
71. Young AT, Tackett MY. Globalization and the decline in labor shares: Exploring the relationship beyond trade and fiscal flows. Eur J Polit Econom. 2017;52:18–35. [Google Scholar]
73. Carvalho VM. From micro to macro via product networks. J Econ Perspect. 2014;28:23–48. [Google Scholar]
74. Saavedra S, Rohr RP, Gilarranz LJ, Bascompte J. How structurally stable are global socioeconomic systems? J R Soc Interf. 2014;eleven:20140693. [PMC free article] [PubMed] [Google Scholar]
75. Battiston South, et al. Complication theory and financial regulation. Science. 2016;351:818–819. [PubMed] [Google Scholar]
76. Hidalgo CA, Hausmann R. The edifice blocks of economic complexity. Proc Natl Acad Sci Us. 2009;106:10570–10575. [PMC gratis article] [PubMed] [Google Scholar]
77. Hidalgo CA, Klinger B, Barabási A-L, Hausmann R. The product space weather condition the development of nations. Science. 2007;317:482–487. [PubMed] [Google Scholar]
78. Hausmann R, Hidalgo CA. The network construction of economic output. J Econ Growth. 2011;sixteen:309–342. [Google Scholar]
79. Neffke F, Svensson Henning M. Revealed relatedness: Mapping manufacture space. Pap Evol Econ Geogr. 2008;8:19. [Google Scholar]
81. Muneepeerakul R, Lobo J, Shutters ST, Goméz-Liévano A, Qubbaj MR. Urban economies and occupation space: Tin they become there from here? PLoS 1. 2013;viii:e73676. [PMC gratis commodity] [PubMed] [Google Scholar]
82. Shutters ST, Muneepeerakul R, Lobo J. Constrained pathways to a artistic urban economy. Urban Stud. 2016;53:3439–3454. [Google Scholar]
83. Shutters ST, Muneepeerakul R, Lobo J. Quantifying urban economic resilience through labour strength interdependence. Palgrave Commun. 2015;1:15010. [Google Scholar]
84. Acemoglu D, Garcia-Jimeno C, Robinson JA. State chapters and economic evolution: A network arroyo. Am Econ Rev. 2015;105:2364–2409. [Google Scholar]
85. Morone F, Del Ferraro G, Makse HA. The k-core as a predictor of structural plummet in mutualistic ecosystems. Nat Phys. 2018;15:95–102. [Google Scholar]
86. Mullainathan S, Spiess J. Automobile learning: An applied econometric approach. J Econ Perspect. 2017;31:87–106. [Google Scholar]
87. Autor D. 2019. Work of the past, work of the future. NBER Working Paper 25588 (National Bureau of Economical Research, Cambridge, MA)
88. Shutters ST, Muneepeerakul R, Lobo J. How difficult is it for urban economies to get 'greenish'? Environ Plann B: Plann Des. 2016;43:198–209. [Google Scholar]
Articles from Proceedings of the National Academy of Sciences of the United states of America are provided here courtesy of National University of Sciences
mccorkindalelosin1937.blogspot.com
Source: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6452673/
0 Response to "Artificial Intelligence and Job Loss Peer Reviewed Journals"
Post a Comment