WORK IN PROGRESS
I'm going to research and explore the topic of unemployment and automation. Questions to be answered:
- How has automation affected unemployment in the past?
- Is automation responsible for current unemployment rate?
- How will automation, and more specifically narrow AI, affect unemployment in the upcoming decades (2015-2035)?
On this page I'll list resources I will use for research, and various tidbits of information I might find relevant. Most of the research will be presented via wiki pages.
Resources
The Robots, AI, and Unemployment Anti-FAQ (by [Eliezer Yudkowsky][2])
Crossposted: http://zanaduu3.appspot.com/pages/3b
Related:
http://www.gwern.net/Mistakes (by Gwern)
http://marginalrevolution.com/marginalrevolution/2013/08/eliezer-yudkowsky-asks-about-automation.html
Comments
Curtis SerVaas
"Humans Need Not Apply" is a good youtube video on this: https://www.youtube.com/watch?v=7Pq-S557XQU
I'd like to see this discussed industry 1. by industry, 2. by technology/science.
(Technology: Computer vision) could impact (Industries: automated farming, automated transportation, etc).
(Industry: Retail/Fast-Food) could be impacted by (Technologies: Using ipads) resulting in (X number of jobs being replaced). The cost of a worker is Y, the cost of an ipad is Z. Consumer polls should L% of consumers thought ipads at fast-food restaurants were creepy.
Honestly, I don't understand why there are still people at the register at fast-food restaurants (as opposed to i-pads). In some restaurants, this is already the case. Is it just a matter of time? If so, why hasn't it happened yet?
Eliezer Yudkowsky
The key question is not whether particular industries get automated. They will be. But so was weaving and farming (3% agricultural employment in First World countries that export food, vs. 98% agricultural employment in the old days) and that didn't impact long-term unemployment. So the question is not whether particular jobs get automated, but whether employment as a whole will marginally drop as a result of marginally better automation given continuation of other current trends, potentially including incompetent central banks that allow aggregate demand slumps and so on.