home | @_patrickoden | messages: p at this website
Skepticism of superhuman artificial intelligence can be compatible with a belief that a large part of our workforce (globally) will eventually become automated. Skepticism of superhuman AI is the null hypothesis. Proponents of the belief in possibility of superhuman AI have a heavy lift at this piont.
Meanwhile, it is a mistake to confuse skepticism of superhuman AI with skepticism of advanced AI and of advanced robotics and of a highly-automated global workforce. That there may still work for humans in the future that robotics and machine intelligence cannot do as well does not mean that there will be enough work for every person to do.
For example: Stipulate that robotics will be able to take over even merely 50 percent of the work that currently employs humans. In 2012 numbers, that would mean the workforce drops from 3 billion people to 1.5 billion people, while the number of unemployed soars from 200 million to 1.7 billion people.
Perhaps this alternate scenario is not the end of humanity that the superhuman AI theory proponents fear, but maybe it sets us on the path toward a different disaster of our own making. The consequences of that many unemployed people around the globe would be devastating, if not species-threatening.
Be skeptical of claims based on a future full of superhuman machine intelligence. Be cautious of writing off all warnings based on a future full of advanced machine intelligence.
The robots coming for even just a lot of our jobs could put us in a perilous position.
 Assume that this 50 percent includes gains from new opportunities from increases in productivity and innovation that come with the robotics/machine intelligence revolution.
01:57 Monday, 19 March 2018