Comment 1 in 3; no; how about 3 in 5! (Score 0) 405
If automation doesn't kill at least 50% by 2025 I'll have my robot eat its hat.
What hasn't happened yet is the creation of the IBM PC of robots. There have been a few cracks at it such as the PR2 but I see that as more of a Sinclair than PC. I want a whitebox robot that I can then glue bits on through a PCI type interface equivalent and make it a factory robot, a hospital robot, or an agricultural robot. For instance I was looking at a machine that was making pretzels and someone had called it a robot. I would have called it a slightly adaptable pretzel making machine.
I have two rules of thumb for what I call a robot: One is that it adapts to its environment somewhat; for instance a garbage picking up machine that looked for garbage, picked up garbage, and did other things such as bringing back a full load and dumping it would meet rule one. Rule two(the lesser rule) is that at the core of the machine can be adapted to something else. So the garbage picking up robot could have some bits switched and it could be a mowing robot or a snow removal robot. I am not saying that the actual garbage robot would be swapable, but that the factory that makes them would be at least adapting a central common core.
So a roomba very much meets rule one but is mostly failing on rule two.
And my rules also apply to software that eliminates a job. I suspect that the software that replaces a call center worker will end up being related to the software that replaces a medical doctor on diagnoses. So adapts to is limited environment, and has a common core.
But when robot designers are working with tools that meet both of my rules then the robot revolution will take off and the job losses will be astronomical. Basically any fairly repetitive job that follows a simple set of logical rules is doomed. This describes many many jobs ranging from building cleaners to medical doctors. Oddly enough some lower skilled jobs will require humans for a very long time. Car repair would be a good example. Often when a car breaks in some way things can be disrupted. So that a simple repetitive routine won't work. Things need to be pounded, pried, and even torched to even get things apart. But computers will assist with such a job by helping to diagnose. If robots are going to damage the car repair profession at all it will be by the robotic assembly of cars resulting in more reliable cars and fewer accidents by robotically driven cars.
What hasn't happened yet is the creation of the IBM PC of robots. There have been a few cracks at it such as the PR2 but I see that as more of a Sinclair than PC. I want a whitebox robot that I can then glue bits on through a PCI type interface equivalent and make it a factory robot, a hospital robot, or an agricultural robot. For instance I was looking at a machine that was making pretzels and someone had called it a robot. I would have called it a slightly adaptable pretzel making machine.
I have two rules of thumb for what I call a robot: One is that it adapts to its environment somewhat; for instance a garbage picking up machine that looked for garbage, picked up garbage, and did other things such as bringing back a full load and dumping it would meet rule one. Rule two(the lesser rule) is that at the core of the machine can be adapted to something else. So the garbage picking up robot could have some bits switched and it could be a mowing robot or a snow removal robot. I am not saying that the actual garbage robot would be swapable, but that the factory that makes them would be at least adapting a central common core.
So a roomba very much meets rule one but is mostly failing on rule two.
And my rules also apply to software that eliminates a job. I suspect that the software that replaces a call center worker will end up being related to the software that replaces a medical doctor on diagnoses. So adapts to is limited environment, and has a common core.
But when robot designers are working with tools that meet both of my rules then the robot revolution will take off and the job losses will be astronomical. Basically any fairly repetitive job that follows a simple set of logical rules is doomed. This describes many many jobs ranging from building cleaners to medical doctors. Oddly enough some lower skilled jobs will require humans for a very long time. Car repair would be a good example. Often when a car breaks in some way things can be disrupted. So that a simple repetitive routine won't work. Things need to be pounded, pried, and even torched to even get things apart. But computers will assist with such a job by helping to diagnose. If robots are going to damage the car repair profession at all it will be by the robotic assembly of cars resulting in more reliable cars and fewer accidents by robotically driven cars.