“So, I was reading about some tech people at the U. of Michigan who made a model to provide robots with “a visual search strategy that can teach them to look for a coffee pot nearby it they’re already in sight of a refrigerator” among other search patterns.* The idea is that robots need to learn how to find things in a complex environment the way you do. Of course, you’re the model for the model since you never misplace your keys.”
“Well, I do misplace them sometimes. I’m just glad I don’t have to be aware of some multimillion dollar Stradivarius. As inevitable as a broken bow string, I’d forget where I put it, maybe even before I had to play Vivaldi’s Four Seasons. I’m always looking for something. What’s that they say? Senior moment? Absent-minded? Too busy for my own good?”
“This effort to make human-like robots is getting pretty intense, competitions and prizes and all, and grants galore. Nothing like having money to pursue your favorite tech toy. But seriously, the difficulty of making a human-like robot explains how perfectly faulty we are.”
“What? Perfectly faul….”
“Yes, if we were perfect, then we would be, if we use the medieval meaning of the word, ‘complete.’ And that which is complete doesn’t change. It is what it is. No, here we humans are, struggling to understand not only other humans but also ourselves. That makes modeling a human pretty difficult. Incompleteness is part of what we are. We’re so darn variable, so unpredictable, so, how should I say it? Oh! Yes, incomplete.
“Maybe we’ll be able to invent a robot that can find a coffee pot in a kitchen filled with all those other appliances, cookbooks, and utensils. That would be good, I guess. And if it could find the coffee pot unfailingly, then wow! That would be sensational. The next task, of course, is to see whether the robot could take a vacation without going the first 10 miles without asking, ‘Did I turn off the coffee pot?’ Such is the real and incomplete human’s nature. I think I would believe in the human-like nature of the robot if it couldn’t find the coffee pot or remember whether or not it turned off the pot before leaving the house.”
U. of Michigan. 19 Jun 2020. Model helps robots think more like humans when searching for objects. TechXplore online at https://techxplore.com/news/2020-06-robots-humans.html Accessed June 21, 2020.
“Well, I do misplace them sometimes. I’m just glad I don’t have to be aware of some multimillion dollar Stradivarius. As inevitable as a broken bow string, I’d forget where I put it, maybe even before I had to play Vivaldi’s Four Seasons. I’m always looking for something. What’s that they say? Senior moment? Absent-minded? Too busy for my own good?”
“This effort to make human-like robots is getting pretty intense, competitions and prizes and all, and grants galore. Nothing like having money to pursue your favorite tech toy. But seriously, the difficulty of making a human-like robot explains how perfectly faulty we are.”
“What? Perfectly faul….”
“Yes, if we were perfect, then we would be, if we use the medieval meaning of the word, ‘complete.’ And that which is complete doesn’t change. It is what it is. No, here we humans are, struggling to understand not only other humans but also ourselves. That makes modeling a human pretty difficult. Incompleteness is part of what we are. We’re so darn variable, so unpredictable, so, how should I say it? Oh! Yes, incomplete.
“Maybe we’ll be able to invent a robot that can find a coffee pot in a kitchen filled with all those other appliances, cookbooks, and utensils. That would be good, I guess. And if it could find the coffee pot unfailingly, then wow! That would be sensational. The next task, of course, is to see whether the robot could take a vacation without going the first 10 miles without asking, ‘Did I turn off the coffee pot?’ Such is the real and incomplete human’s nature. I think I would believe in the human-like nature of the robot if it couldn’t find the coffee pot or remember whether or not it turned off the pot before leaving the house.”
U. of Michigan. 19 Jun 2020. Model helps robots think more like humans when searching for objects. TechXplore online at https://techxplore.com/news/2020-06-robots-humans.html Accessed June 21, 2020.
RSS Feed