The opportunities and challenges that artificial intelligence (AI) and automation are creating for the labor market are gaining increasing attention in both policy circles and society at large. The creation of the American Workforce Policy Advisory Board, whose members are expected to be announced early this year, is intended to “prepare Americans for the 21st century economy and the emerging industries of the future” in light of the rapid changes that “technology, automation, and artificial intelligence” are generating. These changes are expected to arrive soon. Kai-Fu Lee, a Chinese venture capitalist, recently predicted on 60 Minutes that AI could displace up to 40% of the world’s jobs within the next 15 years.
It therefore should come as no surprise that the State of California—home of Silicon Valley and the fifth-largest economy in the world—is starting to examine the impact of AI, automation, and the resultant Technology-Induced Displacement of Employees (TIDE) more closely. On November 27, 2018, California’s Little Hoover Commission released a report titled Artificial Intelligence: A Roadmap of California. The Little Hoover Commission is a well-respected, non-partisan, and independent state oversight agency with a mandate to help improve the effectiveness and efficiency of state government. The Commission described the Roadmap as “not merely a Commission report,” but also “a call for action”:
Other states, cities and countries are surging forward with strategic plans to harness the power of artificial intelligence in ways that will improve their economy, public health and safety, jobs and environment. The race to develop and use AI for good is more akin to a marathon than a sprint. It is fast paced and highly competitive, and one that California should be leading but is not.
Much of the report focuses on areas of potential AI application by the public sector. This Insight will focus, however, on another major theme of the report—the massive impact that AI and other sources of automation will have on the labor market both in California and around the world. The portions of the report discussing the anticipated labor-market changes that AI and automation will bring, and the recommendations that the Roadmap makes with respect to those changes, are in full accord with recent publication from the Emma Coalition, a non-partisan group of employers and industry associations first announced by Littler and Prime Policy Group last year. While the Roadmap was written for the State of California, its information and recommendations hold lessons for government institutions as well as individual employers across the country.
AI, Automation, and the Labor Market
The Roadmap defines automation as “the replacement of repetitive human physical or mental tasks with purpose-designed technology.” This development will have various effects on both the nature of many current jobs and on the labor market as a whole:
[A]utomation and AI continually create, as well as eliminate, tasks and jobs. Both automation and AI will free up human time and capital, reduce product and service cost, create wealth and are themselves new tools that become cheaply or freely available to use in new ways. The new tasks and capabilities that tomorrow’s workers will gain by using AI-enabled products, services and platforms in the workplace, the new jobs AI will create and the new forms of training that AI will enable in the workforce, are hard to predict.
AI generates understandable concern about the future of work, but the Commission noted a number of clear benefits that AI will bring to the workplace and workforce. The automation of dull and dangerous tasks will reduce the incidence of workplace injuries, both improving working conditions and reducing the cost of workers’ compensation programs. AI can also be leveraged to improve access to education and training for current and future workers alike, although the greatest benefits from the use of AI in educational settings will be as a supplement to, rather than a replacement for, human-driven instruction.
That dynamic—AI enhancing rather than replacing humans in the performance of key tasks—may serve as a useful paradigm for the implementation of AI in the workplace more generally. The Roadmap indicated that AI could make many jobs easier “by performing mundane tasks,” which would free up humans to engage in “more interpersonal and creative activities” at work and would “likely promote higher productivity and individual job satisfaction.” Accordingly, for many workers, “the nature of their job likely will change through partial, rather than complete, automation of activities, or by working alongside robots and smart machines to complete critical job tasks.”
Indeed, one witness who testified before the Commission suggested that these broader workplace effects will constitute the main impact of automation for most workers:
All of the attention to date has been on automation. But new technologies have many other direct effects on tasks–deskilling or upskilling existing ones, creating new ones–as well as a slew of indirect effects, such as enabling outsourcing and changing the job matching process. These are all important effects for policymakers to understand. In particular, it is likely that incremental changes in task content and skill requirements will affect more workers in the coming years than large-scale automation events.
Upskilling, Reskilling, and Training
Regardless of whether automation changes the nature of a worker’s job or eliminates it altogether, the upshot of automation for most workers is that they will need to learn new skills—perhaps quite quickly—to remain an active participant in the workforce.
The Roadmap aligns with the observations made in the Emma Coalition’s report, The Future is Now: Workforce Opportunities and the Coming TIDE.1 The Commission suggests that many workers will face upskilling or deskilling in their current job, depending on whether it is higher-skill or lower-skill tasks that succumb to automation. The automation of lower-skill and more mundane tasks should theoretically allow a worker to focus on higher-skill tasks, but that adjustment may require the worker to acquire the necessary skills through additional education or formal vocational training (upskilling). Conversely, if it is higher-skill tasks that are automated, the worker may face deskilling, a process that tends to drive down wages and diminish the value of the worker’s existing skill set.2
Both upskilling and deskilling require affected workers to learn additional skills to maintain their financial security. As stated in a 2017 report by McKenzie Global Institute, which the Commission quoted in the Roadmap:
Some workers will have to be retrained to work alongside AI directed machines, while others will have to be redeployed within the company or elsewhere in the economy; businesses have a vital role to play in aiding these transitions. This will require changes in skills, mindsets and culture as we transition into a world where ‘coworkers’ include machines as well as other people.3
This process of reskilling the current workforce is the central workplace challenge that the Roadmap identifies. One witness summarized the challenge thusly:
The question is: can we actually transition people from what they’re doing now to what they will need to do in the future? This will affect all of us. All of us will have a significant number of activities that machines will be able to do. Then the question is not about mass unemployment but mass redeployment.
This mass redeployment will require considerable investments in technical and vocational education and training. The question remains, however: who will pay for those essential investments? The Emma Coalition’s premise is that employers can do a great deal to prepare themselves and their workers for the TIDE, but that a truly robust response will require engagement and resources from the public sector as well. Little Hoover’s Roadmap focused on the crucial latter component of TIDE preparation.
The Commission reported that there is a general consensus that policymakers, like employers, can positively influence the future of employment—but only if they take appropriate action, and soon. To put the issue into focus for California policymakers, the Commission used the following series of questions for policymakers to consider when assessing how to “build a human infrastructure” that can withstand the disruptive labor-market effects of automation:
- What is the scale and changing nature of work due to AI? How many jobs will be gained and lost by type?
- What will be the pace of the changes in jobs due to AI?
- What types of jobs will change and when will they be most affected?
- What are the education and training requirements for major current job categories and how well do existing institutions of government meet those needs?
- What is the gap workers will face between their current skills and new job requirements over the next 20 years due to the changing nature of work arising from AI?
- Should California establish a lifelong learning account? How should it be structured, who should run such a program, how much will it cost and how should it be financed?
- What changes, if any, should be made to the unemployment insurance program to integrate into a method for securing improved worker training?
- What changes should be made with respect to better preparing students for AI in: (1) the University of California system, (2) the California State University system, and (3) California Community College system? Are any changes necessary in curriculum, financing tuition, the duration of enrollment or the types of degrees or certificates offered?
- What role should the private sector play in worker training? What role should state policy play in encouraging or subsidizing such training? How can the private sector assure training in more remote, rural or hard-to-reach urban communities? What role should apprenticeship programs play? What role should organized labor play in education and training?
- What changes should be made by state government to prepare the current and future workforce for AI? What role should state government unions play?
Other Governmental Responses to AI Around the World
Perhaps to suggest possible answers to these questions, the Commission surveyed some of the steps other governments—both in the United States and globally—have taken to help prepare their own workforces for these transformative changes. The Roadmap noted the early efforts of the U.S. federal government in the AI policy sphere, including the establishment of the Select Committee on Artificial Intelligence, the formation of the bipartisan Congressional Caucus on Artificial Intelligence, and the House Government Reform Subcommittee on Information Technology’s September 2018 report, Rise of the Machines: Artificial Intelligence and its Growing Impact on U.S. Policy. The Commission specifically quoted the House report’s statement that the “United States cannot maintain its global leadership in AI absent political leadership from Congress and the Executive Branch.”4
The Commission went on to cite various efforts occurring at the state and local level. These efforts include future workforce task forces that were recently created in Indiana and Washington, and a Massachusetts Commission on Digital Innovation and Lifelong Learning at the Massachusetts Institute of Technology. The latter commission has a mandate to “examine how the state can expand affordable, high-quality, employer-aligned education, training and lifelong learning opportunities in an era of technological change.” The Roadmap also noted the formation of a working group in Hawaii to examine whether the state should establish a universal basic income for its citizens, an idea that has gained increasing currency in both the tech world and the population at large as a potential policy response to the workforce disruptions of automation.
The Commission then turned its attention to the somewhat more intensive responses that other developed- and emerging-market governments have undertaken. These steps include:
- A multi-pronged European Commission strategy first announced in 2018 to boost investment and set guidelines for AI within the EU.
- Russia’s military-focused investments in AI.
- China’s “plan to lead the world in artificial intelligence technology by committing to invest capital resources into AI research and development,” as well as its inclusion of “AI-related courses in primary and secondary education.”
- Well-funded national AI strategies in France, South Korea, and Japan.
Most of these national initiatives involve substantial investments of state resources, usually paired with extensive private-sector partnerships.
After publication of the Roadmap, Germany announced a national AI initiative, “Artificial Intelligence (AI) Made in Germany.”5 Like initiatives in other countries, the German initiative is to be backed by substantial investments from and cooperation between government institutions and private-sector employers.
Throughout the report, the Commission expressed its view that “California’s state government has not kept pace with national, state and even city governments on organizing efforts to address the implications of AI.” The Commission urged the California legislature and agencies to be more proactive in preparing for the impact of AI, noting that:
A general consensus exists among experts that lawmakers, educational institutions and private sector policymakers must act quickly to influence not only the impact of AI on employment, but to propel California as a leader in the transition of the labor force as a result of AI and accelerate statewide economic prosperity emanating from AI.
The Commission made a number of specific recommendations, including several focused specifically on AI’s anticipated impacts on the workplace and workforce:
Recommendation 6: The state of California—in recognition of the central importance of relevant, useful and unbiased information—should implement changes to improve the collection of data on jobs and skills for future workers.
Recommendation 9: The Governor and Legislature should improve the data collection related to at-risk jobs. In particular, consideration should be given to requiring the Employment Development Department (EDD) to develop an addendum survey to the Bureau of Labor Statistics (BLS)’s survey that asks questions about the impact on jobs of automation and, specifically, AI technology. To maximize the collection of data related to at-risk jobs, there should be improved evaluation of data that, at a minimum, identifies how the impact of AI technology and applications will differ by region, county, employment sector, socio-economic group, educational attainment, age group, gender and other appropriate characteristics.
On the topic of data collection, the Commission specifically pointed to a system developed by the California Community College Centers of Excellence (COE) for Labor Market Research, which “conducts special projects to better understand the workforce needs in evolving and emerging industries. By conducting point-in-time surveys and talking to employers in controlled ways, the COE has more accurate real time information of job trends.” The Commission expressed its view that the COE’s data collection would be an improvement on the data currently used by state agencies, but that “[e]xpanding the role of COE would require legislative approval.”
The Commission also made recommendations on the topic of workforce training:
Recommendation 5: State government, including the executive branch and independent entities such as local school districts, government funded or subsidized training providers, the U.C. system and the California State University (CSU) system, as well as regional workforce development entities, should create and implement a comprehensive strategic plan to better assure that future workers are prepared for the new skills required for the jobs in a rapidly evolving workplace focused on the application of technology including AI.
Recommendation 10: The Governor and Legislature should promote apprenticeships and other training opportunities for employees in private industry, and state and local governments whose jobs and/or classifications may be displaced or transformed by AI technologies and applications.
To help achieve these workforce goals, the Commission stated:
The state can partner with non-profit organizations to establish better online portals for access to skill assessments, training resources and job search tools, such as the Council for Adult and Experiential Learning (CAEL). California also can actively notify workers once they apply for unemployment insurance that they qualify for unemployment insurance benefits if they participate in approved training. Provided it could expediently implement workforce training and retraining policies, the state should prioritize collaborations with stakeholders.
The need for U.S. state governments to be more proactive in addressing the workforce disruptions from automation was a major topic of discussion at a roundtable that Littler hosted in San Francisco on November 12, 2018. As noted in a WPI report on the roundtable, “[s]tates have traditionally been laboratories for experimentation for labor-market policies.”6 While state engagement in TIDE preparation is not sufficient to help prepare the American workforce for the transformative changes on the horizon, state-level investments and policy innovations will be essential components of an effective national response.
If there was a single overarching takeaway from the Roadmap on the topic of the TIDE and workforce training, it was that policymakers still have an opportunity to positively influence the future of employment—but only if they take appropriate action and soon. The same is true of America’s employers in both the public and private sectors, who remain best-positioned to identify the skills that employers will require in the future and establish training programs to provide workers with the necessary skills.
These issues will only increase in importance and in public consciousness over the coming years. Littler, WPI, and the Emma Coalition stand ready to assist employers in navigating these transformative changes.