As the dust settles on GM’s announcement to lay off nearly 15,000 workers, several things are becoming apparent to people who watch AI’s impact on jobs:
- AI’s threat to labor can come not just from the automation of human tasks, but also from sudden, drastic shifts in corporate investment toward AI.
- The modus operandi of lay off & hire fresh versus upskill & transition is alive and well, at scale, and industries beyond manufacturing may be just as susceptible.
- “AI investment” may be the new executive cover story for downsizing; GM’s shares initially jumped on the announcement – with the financial press speculating that the shift to autonomous vehicles is a fundamental positive.
- Federal policies lack a cohesive understanding of AI’s role in the future of work.
Optimized Workforce holds no preconceptions regarding AI’s effect on labor. We do not expect it to drive a social utopia nor fuel a universal basic income – yet data from our “2018 AI Preparedness Survey” indicate we are not at the cliff of an unemployment apocalypse either. However, to date, we have studied task-level advances in technology; our data set cannot forecast brash executive-level budget reallocation. GM alone can’t ruin a healthy labor market – even one that seems poised to methodically embrace AI in the short term. But GM can start a trend.
Nothing in Mary Barra’s public statement nor the thus-reported buyout plans offered to GM employees contains offers for retraining or job transitions. This may be happening behind the scenes, but the public perception here is, “Line workers, you’re out. Tech workers, you’re in.” This could be a significant oversight on GM’s part. As part of our “2018 AI Preparedness Survey,” Optimized Workforce tested respondents on factors that we believe indicate a worker’s likelihood to succeed alongside AI. As an industry, manufacturing performed slightly better than the U.S. workforce at large on these tests. If GM’s move is oversight, it can also set a bad precedent for the U.S. economy at large.
Back in the Great Recession, companies slashed workforces out of necessity – but after a period of downsizing, many of us who were employed during that era suspected something else was afoot. Two years after the recession’s official end (June 2009), the S&P was up 45% and corporate profits were increasing at an 18% annual clip, but employers were not rehiring. The people still employed seemed to be working harder, longer. Our findings here are anecdotal, but many on the Optimized Workforce board remember full-time job reqs being regularly denied, even when working conditions demanded it. The specter of recession permitted executives to delay hiring. Will the specter of budget allocation for AI investment create a similar phenomenon?
Lastly, the federal response to GM’s moves should raise concern. Threats to pull subsidies from GM or subject it to a very pointed set of tariffs because it has chosen to innovate will not stop innovation in the long run. Applying a knee-jerk punishment is no more useful than a campaign promise to bring back U.S. manufacturing jobs – when no one on either side of the political aisle can do that through policy alone. The U.S. march toward an AI-driven economy, be it a dawdle or a sprint, is going to happen. In a global economy it has to happen. The federal government should be studying this evolution and crafting policies to aid in worker transitions and create sustainable governance for employers who adopt the technology, it should not try to put a genie back in a bottle.
This post is not intended as a critique of GM’s decision to restructure. If the product lines that the plants support underperform, and if the long-term vision for growth at GM does not include these products, GM has no choice but to make these moves. But the way they’re making them, and the policy reactions they are causing, might be the first widely reported case study of the U.S. workforce’s shift toward AI. It’s not a soothing story.
Image used under a creative commons license, courtesy https://www.ncpedia.org