Philip Ball, "The robot within," New Scientist, March 15, 2003
Bill Joy, "Genetics, nanotechnology, robotics pose danger to society," The Sunday Patriot-News Harrisburg, July 23, 2000
Robert A. Freitas Jr., "Some Limits to Global Ecophagy by Biovorous Nanoreplicators, with Public Policy Recommendations," Foresight Institute, April 2000
How could this happen? It's one of the inherent dangers posed by nanotechnology technology at the scale of a billionth of a meter (or about the width of an atom or two). Nanotechnologists envision (indeed, are working toward) a world in which nano-scale assemblers can be programmed to build anything cars, cardigans, carving knives, you name it by manipulating individual atoms (taken from, say, a nearby load of scrap metal or sand). To make this efficient, trillions of these assemblers are required, so the first task is to make assemblers that can forge copies of themselves, so-called replicating assemblers or nanoreplicators.
In a production environment, these replicating assemblers would operate under strict supervision that shuts down the replication process when enough machines have been made. But what would happen if just one of these unimaginably small replicators escaped from this controlled scenario? Without other safeguards in place, possibly the only thing that would stop the machine from making copies of itself is running out of whatever raw materials it needs to perform its work.
Now suppose, however, that the machine is biovorous (literally, life-eating) meaning that it uses organic matter as its raw material. In this case, theoretically, the machine and its offspring would, under certain circumstances, continue replicating until all the world's plant and animal life was consumed: global ecophagy.
This is also known, more vividly, as the gray goo problem, since all that would be left of the world is an undifferentiated mass of nanoreplicators. The problem was first expounded by K. Eric Drexler in his book Engines of Creation:
|
blue goo
climate porn
eco-tech
global dimming
normal accident
precautionary principle
revenge effect


