The Future of Work in the Age of AI: Displacement or Risk-Shifting?

Artificial Intelligence and Innovation and Economic Growth

Article Snapshot

Author(s)

Karen Levy and Pegah Moradi

Source

The Oxford Handbook of Ethics of AI, Markus D. Dubber, Frank Pasquale, and Sunit Das, eds., pp. 27-52, 2020.

Summary

Artificial intelligence-based systems (AI) are altering the conditions and quality of work. Employers use AI to shift risks to low-wage workers, adopting irregular schedules or systems that force workers to adopt a strenuous pace.

Policy Relevance

Policymakers should adopt measures that improve the conditions and quality of work.

Main Points

  • One study predicts that 47 percent of jobs in the United States are at high risk of being automated, with most being low-wage occupations; such forecasts ask whether automation is technically feasible, but neglect social factors affecting automation, such as the political power of white workers.
     
  • AI may change the task composition of jobs, redefining the occupation rather than replacing it; the growth of online sales will eliminate retail stores, but might create different jobs elsewhere in the supply chain.
     
  • Unemployment might result as tasks are altered to make them easier for AI to manage; AI systems struggle to place items into properly sized boxes, so online sellers are developing systems that use AI to build boxes around items.
     
  • Whether AI will displace workers is unclear, but AI is already affecting the conditions and quality of work; firms are using AI-based systems to shift risks from themselves to workers.
     
  • With algorithmic scheduling, employers use sensors to measure shifts in demand, resulting in fluctuating "just-in-time" schedules that introduce instability into the lives of low-wage workers.
     
  • Firms pinpoint “indispensable” tasks for which compensation must be paid under the Fair Labor Standards Act, but deny compensation for related tasks, such as donning protective gear.
     
  • AI can be used to shift the risk of misconduct by employees (such as theft) from the employer to the employee.
     
    • Firms use AI to monitor worker’s behavior and flag unusual activity.
       
    • Some predictive systems for preventing loss might be biased against disadvantaged workers.
       
  • Firms use AI to detect and penalize short lapses in employee productivity, resulting in extremely strenuous working conditions.
     
  • Many policy proposals seek to address the problem of workers displaced by AI-based systems; policymakers should also consider protection for workers who retain jobs, such as laws that require employers to announce schedules in advance.
     

Get The Article

Find the full article online

Search for Full Article

Share