Event Recap: The Preparing the Workforce of the Future Conference
On October 7th, the Block Center convened workforce economists and education technology experts on our campus for discussions on fast-growing skills and jobs, and innovative methods that educators, employers, and policymakers can use to prepare workers for these shifting career opportunities.
We packed a great deal of research-based discussion into the day, starting with a presentation by Lisa Kahn (University of Rochester) showing preliminary evidence that when worker training programs at firms are co-funded by the government, they lead to more job growth at the firm. This is in contrast to more traditional worker training programs which typically only target unemployed workers and do not involve the private sector. This suggests public-private partnerships in workforce training programs can play a role as the need for mid-career upskilling aligned with technology advancements becomes more acute. While Kahn focused on one driver of job postings, Sonny Tambe (University of Pennsylvania) presented evidence on who job postings attract. His study on tech industry job openings posted on AngelList revealed that when postings advertise a remote setup, they attract, on average, more diverse applicants than do equivalent job postings that advertise an in-person setup.
The participating education technology scholars offered evidence of promising solutions — many involving Artificial Intelligence (AI) tutors—that can help prepare students for new job opportunities created by innovating firms and the growth of remote workplaces. These solutions are crucial in the face of pandemic-induced learning loss, a widespread teacher shortage in the U.S., as well as a growing acceptance of remote education offerings. Danielle McNamara (Arizona State University) gave a keynote talk that touched on ASU’s tutoring program Writing Pal, which offer gamified reading and writing instruction with feedback enhanced by NLP.
Two economists showed evidence supporting the potential of these technologies using randomized controlled trials, the gold standard in empirical research. Philip Oreopoulous (University of Toronto) evaluated the effect of providing teachers with computer-assisted learning modules, showing this classroom and homework tool improved math test scores for students in grades 3 through 9. Anuj Kumar (University of Florida) showed that personalized math homework enhanced student learning in India, highlighting that technology-enabled education services have promise also in developing country settings. Several Block Center scholars at the frontier of developing and implementing these education technologies presented the work being done by their research groups, including Ken Koedinger, Carolyn Rose, and Majd Sakr, whose Sail() platform offers AI-enabled training for in-demand IT skills.
The public portion of our conference featured an engaging lecture by David Autor (MIT) on how advancing technology is affecting the labor market, with a reminder that the emergence of new job roles has been a pervasive feature of the market: from Airplane designer (first recorded by the Census in 1950) to Wind turbine technician (2010) to Cybersecurity analyst (2018). We wrapped up with a panel moderated by our Block Center visiting scholar Anna Salomons (Utrecht University) on how we can “make AI work for us.” One highlight was Avi Goldfarb’s (University of Toronto) discussion of two ways to think about how AI adoption affects industry and jobs: by either automating steps of existing processes without truly “disrupting” them (Goldfarb conjured an AI dispatcher for a taxi company), versus creating completely new processes (such as an AI tool that allows anyone with a phone and a car to become a taxi driver). While it is relatively easy to predict the jobs created and eliminated by the former, the workforce implications of the latter requires more creativity and forethought, but the gains to society would likely be larger as well.
As an example of another potentially transformative change, Sameena Shah (J.P. Morgan AI Research) argued that AI has the potential to help overcome persistent biases in human decision-making. More generally, all panelists agreed that researchers, policymakers, educators, and firms should not simply be trying to predict what the labor market will look like with advancing AI adoption, but rather aim to steer the development and implementation of these technologies for maximal societal benefit.
Thank you to everyone who joined us in person and via the live stream. For a taste of the action, check out the recording of David Autor’s keynote talk and the cross-sector panel here.
