> Robots They're rather less common or good than Asimov supposed. As with cooking gadgets, it turns out that it's difficult to design something that can navigate a home (including stairs and uneven surfaces), perform tasks, usefully, and not cost a bundle. Especially when competing with low-cost maid services.
He did say they'd be uncommon and not very good. What he didn't forsee is that they would not be humanoid robots.
Roombas are available, and can help clean your house. They aren't all that common, and aren't all that "good" by his metric of what a robot should be able to do, but they do exist, and people really do use them.
Furthermore, UAVs are increasingly common in the military. While some of them, like Predator drones, are remote controlled, some of them can operate fully autonomously for the vast majority of their flight.
And of course, industrial robots are so ubiquitous that they've faded into the background long ago.
So I would say that "uncommon and not very good" (according to his vision of humanoid robots that can replace most manual human tasks) is a fair assessment. He just figured that they would be clumsy humanoid robots, as opposed to more specialized robots that were fairly good at an individual task but can't do anything else.
Calling a Roomba a robot ... is somewhat on par with calling a blade of grass a tree. A Roomba randomly careers across a floor probabalistically providing sufficient coverage to fully vacuum or mop it.
Yes, we've got more intelligent robots, but they're a tad more expensive. Industrial manufacturing robots exist, but the reprogramming costs exceed the manufacture cost of the robots themselves. They make sense in very high-value, generally high-volume production. They're not particularly adaptive, though they can switch between different modes (programs) fairly well.
Flying through empty space is a reasonably simple problem. We've had ICBMs for ages. Navigating through complex terrain (ground-based robots, autonomous vehicles) is a markedly more complex domain space.
Most robots we've got at present are very, very highly specialized. Or incredibly stupid, though good enough for the task at hand.
He did say they'd be uncommon and not very good. What he didn't forsee is that they would not be humanoid robots.
Roombas are available, and can help clean your house. They aren't all that common, and aren't all that "good" by his metric of what a robot should be able to do, but they do exist, and people really do use them.
Furthermore, UAVs are increasingly common in the military. While some of them, like Predator drones, are remote controlled, some of them can operate fully autonomously for the vast majority of their flight.
And of course, industrial robots are so ubiquitous that they've faded into the background long ago.
So I would say that "uncommon and not very good" (according to his vision of humanoid robots that can replace most manual human tasks) is a fair assessment. He just figured that they would be clumsy humanoid robots, as opposed to more specialized robots that were fairly good at an individual task but can't do anything else.