6 ms·

Is this a trick question? Smart enough (along with the rest of my home) to order the cheapest reliable courier to pick up a quantity of fresh vegetables suffic

Is this a trick question? Smart enough (along with the rest of my home) to order the cheapest reliable courier to pick up a quantity of fresh vegetables sufficient for the dietary profile that I need based on my height and weight and dietary restrictions, to order the vegetables from my local grocer for the courier to pick up, to let the courier into my apartment when they arrive and put them in the fridge, and smart enough to get them out of the fridge and make me a smoothie at the time that I've ordered on my smart phone, and then go wash itself in my dishwasher. Why - how dumb do you want your blender to be? Just a spinning blade for you to leave in a grimy corner of your kitchen or stowed away in a cupboard, never to be used? of course, some of this functionality logically overlaps (why should the blender put itself in my dishwasher rather than my dishwasher taking the blender; is the blender ordering my groceries or is my fridge) but you get the idea....

derefr 10y ago
I think there are two seemingly-disjoint perspectives in these discussions, mostly based on whether people have enough time and engage with these appliances as part of a hobby; or whether they are strapped for time and engage with them as necessities. So, I think a clarifying assumption might help: assume that you already have a person who acts in the role of the "glue" for your household—a https://en.wikipedia.org/wiki/Majordomo https://en.wikipedia.org/wiki/Majordomo. Assume that they then have further staff under them, to whom they can delegate tasks—but that they also have the option of purchasing technology to automate the same tasks. Under this assumption, all the tasks in your household will get done, "magically", by combining some degree of technology with some degree of human effort. The question that's really being asked in these discussions of "smart" appliances is: where is the optimum ROI between "fully automated" and "fully done manually by a paid human expert"? For any given task, how should the majordomo best allocate your dollars between paying labor and buying "smarter" equipment? Automation usually lowers the total amount of human labor that must be expended (both by the expert, and by you to earn the money to pay the expert), but it also tends to lower the quality of the result—or "confine" the result to certain strict output products—compared to the results the human expert could achieve. How "smart" can we make a tool without taking away its precision to achieve great results in expert human hands? How "dumb" must a tool be to continue to allow for optimal results to be achieved, rather than mere satisfactory ones? Etc.

derefr 10y ago

I think there are two seemingly-disjoint perspectives in these discussions, mostly based on whether people have enough time and engage with these appliances as part of a hobby; or whether they are strapped for time and engage with them as necessities. So, I think a clarifying assumption might help: assume that you already have a person who acts in the role of the "glue" for your household—a https://en.wikipedia.org/wiki/Majordomo https://en.wikipedia.org/wiki/Majordomo. Assume that they then have further staff under them, to whom they can delegate tasks—but that they also have the option of purchasing technology to automate the same tasks. Under this assumption, all the tasks in your household will get done, "magically", by combining some degree of technology with some degree of human effort. The question that's really being asked in these discussions of "smart" appliances is: where is the optimum ROI between "fully automated" and "fully done manually by a paid human expert"? For any given task, how should the majordomo best allocate your dollars between paying labor and buying "smarter" equipment? Automation usually lowers the total amount of human labor that must be expended (both by the expert, and by you to earn the money to pay the expert), but it also tends to lower the quality of the result—or "confine" the result to certain strict output products—compared to the results the human expert could achieve. How "smart" can we make a tool without taking away its precision to achieve great results in expert human hands? How "dumb" must a tool be to continue to allow for optimal results to be achieved, rather than mere satisfactory ones? Etc.