Opinion polls show major public worries about artificial intelligence (AI) taking jobs and creating financial uncertaintiesfor the next generation. Although those fears are understandable given the rapid pace of technological change, the actual impact is likely to be more gradual; research on job impacts shows widely varying estimates, and there is little agreement over how fast or substantial those employment consequences will be.
We shouldn’t be surprised at these worrisome reactions because Big Tech leaders themselves are warning about dismal job futures or the possibility of “unusually painful” job losses. Given the gravity of the current moment, there is an urgent need for better publicly available data on the digital economy and more effective workforce development programs.
The need for better public data
It is critical to compile and analyze data on the job impact and workforce training needs of the digital economy. One of the reasons there is so much uncertainty regarding the future of work is that U.S. government data collection systems remain mired in the industrial era. Federal agencies are very good at surveying businesses and consumers about factory production, international trade, and monetary flows but are not as effective when it comes to modern digital technology.
As a result, federal agencies lack publicly available data on many aspects of the digital economy, including how quickly firms are incorporating AI into their internal operations, how emerging technologies are affecting hiring and firing, and the way firms are switching to temporary workers or independent contractors.
Yet it is not that those kinds of data do not exist. The digital economy generates an extraordinary amount of very helpful information about the tiniest aspects of people’s online behavior as well as those of tech firms. But almost all of that information is considered proprietary and is thus typically not released to researchers. That means policymakers are flying blind at a time when private companies are compiling digital data without making it available for researchers or government officials to analyze and project the impact.
Improving workforce development
Companies need to get serious about workforce development. If we are to avoid catastrophic job losses, professional development opportunities are required to offer needed skill development and ensure workforce development programs train people for where the economy is going over the next five to 10 years. Education needs to be regarded as lifelong learning, and many individuals will need to continually upgrade their jobs skills. The idea that education ends at college needs to give way to one in which people continually update their skills throughout their working lives.
Several key features are required for workforce development programs. First, job training must be geared to local economic needs. Local development councils that now handle a lot of current training need up-to-date data and projections so that training can be aligned with current and future economic and job demands. Some workforce development programs have poor placement rates and are not providing the proper training for an AI era where algorithms and agents are supplanting human skills. Training personnel need to determine which tasks are being supplanted by AI, which ones are being supplemented, and which skills will continue to be in demand in coming years.
In recent months, some industries have recognized the need for career alignment and have begun guaranteeing jobs at the end of the training program. A few are linked to specific companies, while others are tied into partnerships with labor unions that promise good jobs with good wages and benefits. As an illustration, both Google and Microsoft recently announced major investments in data center construction. Their programs are targeting skilled labor areas where well-trained electricians, plumbers, pipefitters, and the like are in demand.
Data center construction requires many of those individuals, yet they remain in short supply in many different regions. Matching supply and demand is critical for successful workforce development programs. Additionally, skill credentials must be transferable and respected across regions. Meta has a Workforce Academy where the certification moves with the worker. This is a key benefit in many vocational areas as workers need the flexibility to move to regions that may be far away from their training program. Nontransferable credentials have far less value than ones that are mobile and useful in a variety of geographic areas.
Finally, partnerships also matter; one such example is Amazonand Northeastern University’s alliance that continues to offer apprenticeships for entry-level workers. Community colleges are also likely partners in programs requiring professional certification or skills training. The more industries align their credentials with local entities, the more likely local people will advance in the digital economy.
According to some research, early-career individuals are the ones most likely to be affected by job displacement in the short run. Tasks in those areas tend to be routine and repetitive or consist of administrative functions that are easily automated. There are already algorithms and agents that can perform those tasks, so affected workers would benefit from training programs specifically geared to their needs.
More generally, the fundamental question looming over all these discussions is: Who pays for workforce development? Right now, we have a mismatch of programs run by businesses, governments, nonprofits, and higher education. In some cases, the financial burdens rest almost entirely on private individuals, while in other cases there is assistance from government or businesses. Having robust and timely data will help move the needle on workforce advancement at a time of widespread transformation. Government plays a legitimate role in ensuring worker and industry productivity.
Just as taxpayers decades ago stepped up to finance high school, college, and now pre-K programs, adult education has emerged as a vital area of interest for the country as a whole. As we make the transition to a digital economy, we need to ensure that a broad swath of the country is not left behind. More publicly available data will demonstrate the compelling need to move in this direction.
Read the full story by Nicol Turner Lee and Darrell M. West / Brookings
