Difference between revisions of "AI Research"

Revision as of 16:46, 25 January 2020

Future of Employment: How susceptible are jobs to computerization - 2013 - Oxford University

• Author's - Carl Benedikt Frey and Michael A Osborn
• 47 percent of U.S. jobs could be automated within the next decade or two
• occupation-based evaluation
  • evaluated the likelihood that 71 occupations could be automated
  • modeled the impact on 632 additional occupations
• they estimated what was technically possible

Risk of Automation for Jobs in OECD Countries: A Comparative Analysis (pdf) - 2016 - OECD

• task-based approach
• 9 percent of jobs in U.S. were at high risk for automation
  • based on at least 70% of the jobs task could be automated

Will robots really steal our jobs? An international analysis of the potential long term impact of automation - 2017 - PWC

• 38 percent of jobs in the United States were at high risk of automation by the early 2030s.
• stated jobs that could be automated, yet actual losses will be mitigated by regulatory, legal and social dynamics

Harnessing automation for a future that works - 2017 - McKinsey

• 50 percent of work tasks around the world are already automatable
• 30 percent of work activities could be automated by 2030
• only 14 percent of workers would need to change occupations

2030: The Collision of Demographics, Automation and Inequality - Feb 2018 - Bain and Company

• employers will need 20 to 25 percent fewer employees by 2030
• that's equal to 30 to 40 million displaced workers in the United States
• 80 percent of all workers will be affected

Prediction Challenges

• many jobs have a mix of tasks - some can be automated and other cannot
• entirely new occupations
• enablement jobs
• speed of technology deployment - if it can be automated, it may not be done for many reasons
• demand from lowering the cost of goods and services production
• shift of spending to other goods and services if savings
• AI Startups find a whole new way to satisfy a human need