AI is taking over the workplace. Will there be enough jobs left for people?
The fear: Workers of all kinds are on the firing line as large language models, text-to-image generators, and hardware robots match their performance at a lower cost.
Horror stories: Automated systems are performing a wide range of tasks that previously required human labor.
- Voice-enabled language models take orders at fast-food restaurants. Their mechanical counterparts cook fries.
- Large language models write articles for publications including CNET, Gizmodo, publications that share ownership with Sports Illustrated, and outlets associated with the United Kingdom’s Daily Mirror and Express.
- Image generators are producing concept art for game developer Blizzard Entertainment, and a synthetic image appeared on the cover of a book published by Bloomsbury.
- Humanoid robots are moving bins in Amazon warehouses, while mechanical arms that shape sheet metal fabricate parts for airplanes.
Creeping pink slips: Workers are expressing anxiety about their prospects, and researchers believe the labor market is about to experience a seismic shift.
- 24 percent of U.S. workers worry AI will take over their jobs, a May survey by CNBC found.
- Hollywood writers and actors staged a protracted strike partly over concerns that generative AI would devalue their work.
- Investment bank Goldman Sachs predicted that AI could put 300 million full-time jobs at risk.
Facing the fear: Each new wave of technology puts people out of work, and society has a responsibility to provide a safety net and training in new skills for people whose jobs become fully automated. In many cases, though, AI is not likely to replace workers — but workers who know how to use AI are likely to replace workers who don’t.
- The United States Bureau of Labor Statistics identified 11 occupations at risk of being automated — such as language translators and personal financial advisors — and found that 9 of them grew between 2008 and 2018.
- Human jobs tend to involve many tasks, and while AI can do some of them, it’s poorly suited to others. An analysis of AI’s impact on jobs in the United States concluded that, for 80 percent of the workforce, large language models would affect at least 10 percent of tasks. This leaves room for AI to boost the productivity — and perhaps wages and even job security — of human workers.
- Technological advances typically create far more jobs than they destroy. An estimated 60 percent of U.S. jobs in 2018 did not exist in 1940. Looking forward, consider the likely explosion of machine learning engineers, data scientists, MLOps specialists, and roboticists.