3 ms·

I assume you are referring to driverless cars. Those aren't inherently unsafe. Neither are six axis robots. As I mentioned, cobots are well suited for this sort

by drew-y 4y ago

I assume you are referring to driverless cars. Those aren't inherently unsafe. Neither are six axis robots. As I mentioned, cobots are well suited for this sort of situation. Just to reiterate, I am not trying to make some any sort of statement about the country. I'm just expressing my disgust in how this specific incident was responded to and the multiple failures in oversight leading up to it. Mistakes like this happen everywhere.

delusional 4y ago
I'm genuinely curious, because I don't see it. What's the property that makes these robots "inherently unsafe" while driverless cars aren't?
- mdp2021 4y ago
  The poster has not stated that robots would be inherently unsafe. (And if they were judged so, it would also be in consideration that humans also are.) Anyway: the difference is in explicit projectual intention. Some systems are created with a dimension (contextually, safety) as a key performance indicator, some other as just a factor to be sensibly taken into consideration. If you are drawing the project for a plane, a main thing you'll keep in mind is that it will have to stay up; if a toy, safety is required but not part of the function.
  - delusional 4y ago
    > The poster has not stated that robots would be inherently unsafe. (And if they were judged so, it would also be in consideration that humans also are.) But he has. The statements > This is appalling. There is no world in which an industrial robot like that should be operating next to a child. > I assume you are referring to driverless cars. Those aren't inherently unsafe. Neither are six axis robots. As I mentioned, cobots are well suited for this sort of situation. Together form a very easy to interpret implicit argument that "industrial robots" are "inherently unsafe" (since that was used to distinguish them from driverless cars) and that cobots are, by virtue of not being "inherently unsafe" (like "industrial robots"), well suited. Even just the choice of words for "industrial robots" lay bare a moral judgment of using them for entertainment purposes near children. Is there anything less "industrial" about the 500KW motor on a tesla? If you want to claim that he didn't _directly_ and _explicitly_ state that argument outright, then you're right. It doesn't take a lot of reading between the lines to come to the conclusion that he considers there to be some difference. > Anyway: the difference is in explicit projectual intention. [...] So what's the argument here? That driverless cars were made to be safe and this chess playing robot wasn't? With how driverless cars have been pushed onto the roads with over the top promises and negligent safety features, how do you even begin to substantiate that argument?
    - amelius 4y ago
      Cars (period) are not inherently safe to begin with. Perhaps it would help if you applied your argument to good old regular cars and see where it fails.
      - delusional 4y ago
        I'm not trying to make an argument, I'm trying to examine the argument made by someone else. Personally, I believe that neither driverless cars nor "industrial robots" are "inherently unsafe". In fact I think the whole notion of "inherently unsafe" is flawed. An "industrial robot" playing chess with children can be fine if it's engineered well, which this particular robot evidently wasn't. Driverless cars can be fine if they're engineered well, which I would argue has not always been the case.
    - mdp2021 4y ago
      > reading between the lines Don't do it. You have been here for years, you should have noticed: do not read between the lines. Do not overread what other said. Do not attribute intentions. Do not paint ideas that come from you on the expression of others. In the past very few days I had posted about the buffers of undetermined positions¹, and hours before I had a post of mine literally re-translated in ideas that certainly were only in the mind of the "translator", parallel with other interventions that surely had skipped the duly operation of understanding where possible and stopping inventing statements otherwise. It is really to be avoided. You exercise doubt on your interpretations just like you exercise doubts on all your ideas - and more, on your habit of forming ideas. ¹https://news.ycombinator.com/item?id=32205085 https://news.ycombinator.com/item?id=32205085 (and following, for clarity about them "buffers") Contextually: notice that formula, '«like that»', that the poster wrote. > So what's the argument here Yes, the chief point in engineering a "driverless car" is that it does not crash on things. I «substantiate that argument» - which a difference is there between KPIs and constraints - beyond the beginning because nobody told you that the engineers succeed.
- numpad0 4y ago
  I think it’s an inference from domain specific common sense in robotics. “Inherently safe” is a technical term. It’s an ideal form that almost never materializes, so “functional safety” substitutes it and arguments are made as to what risks exists or how they should be mitigated. “Inherently unsafe” as reverse of that means an object cannot be safe no matter what; A bomb next to a person can be an example. Had a driverless car inherently unsafe, it cannot be placed on public road. The original argument therefor means, that it is established that cars and robots can be safe-fied with functional safety standards and measures such that only tolerable risks exist, so they cannot be inherently unsafe, so generally they can be let be on public roads or around people with conditions.

by drew-y 4y ago

delusional 4y ago

I'm genuinely curious, because I don't see it. What's the property that makes these robots "inherently unsafe" while driverless cars aren't?

mdp2021 4y ago
The poster has not stated that robots would be inherently unsafe. (And if they were judged so, it would also be in consideration that humans also are.) Anyway: the difference is in explicit projectual intention. Some systems are created with a dimension (contextually, safety) as a key performance indicator, some other as just a factor to be sensibly taken into consideration. If you are drawing the project for a plane, a main thing you'll keep in mind is that it will have to stay up; if a toy, safety is required but not part of the function.
- delusional 4y ago
  > The poster has not stated that robots would be inherently unsafe. (And if they were judged so, it would also be in consideration that humans also are.) But he has. The statements > This is appalling. There is no world in which an industrial robot like that should be operating next to a child. > I assume you are referring to driverless cars. Those aren't inherently unsafe. Neither are six axis robots. As I mentioned, cobots are well suited for this sort of situation. Together form a very easy to interpret implicit argument that "industrial robots" are "inherently unsafe" (since that was used to distinguish them from driverless cars) and that cobots are, by virtue of not being "inherently unsafe" (like "industrial robots"), well suited. Even just the choice of words for "industrial robots" lay bare a moral judgment of using them for entertainment purposes near children. Is there anything less "industrial" about the 500KW motor on a tesla? If you want to claim that he didn't _directly_ and _explicitly_ state that argument outright, then you're right. It doesn't take a lot of reading between the lines to come to the conclusion that he considers there to be some difference. > Anyway: the difference is in explicit projectual intention. [...] So what's the argument here? That driverless cars were made to be safe and this chess playing robot wasn't? With how driverless cars have been pushed onto the roads with over the top promises and negligent safety features, how do you even begin to substantiate that argument?
  - amelius 4y ago
    Cars (period) are not inherently safe to begin with. Perhaps it would help if you applied your argument to good old regular cars and see where it fails.
    - delusional 4y ago
      I'm not trying to make an argument, I'm trying to examine the argument made by someone else. Personally, I believe that neither driverless cars nor "industrial robots" are "inherently unsafe". In fact I think the whole notion of "inherently unsafe" is flawed. An "industrial robot" playing chess with children can be fine if it's engineered well, which this particular robot evidently wasn't. Driverless cars can be fine if they're engineered well, which I would argue has not always been the case.
  - mdp2021 4y ago
    > reading between the lines Don't do it. You have been here for years, you should have noticed: do not read between the lines. Do not overread what other said. Do not attribute intentions. Do not paint ideas that come from you on the expression of others. In the past very few days I had posted about the buffers of undetermined positions¹, and hours before I had a post of mine literally re-translated in ideas that certainly were only in the mind of the "translator", parallel with other interventions that surely had skipped the duly operation of understanding where possible and stopping inventing statements otherwise. It is really to be avoided. You exercise doubt on your interpretations just like you exercise doubts on all your ideas - and more, on your habit of forming ideas. ¹https://news.ycombinator.com/item?id=32205085 https://news.ycombinator.com/item?id=32205085 (and following, for clarity about them "buffers") Contextually: notice that formula, '«like that»', that the poster wrote. > So what's the argument here Yes, the chief point in engineering a "driverless car" is that it does not crash on things. I «substantiate that argument» - which a difference is there between KPIs and constraints - beyond the beginning because nobody told you that the engineers succeed.
numpad0 4y ago
I think it’s an inference from domain specific common sense in robotics. “Inherently safe” is a technical term. It’s an ideal form that almost never materializes, so “functional safety” substitutes it and arguments are made as to what risks exists or how they should be mitigated. “Inherently unsafe” as reverse of that means an object cannot be safe no matter what; A bomb next to a person can be an example. Had a driverless car inherently unsafe, it cannot be placed on public road. The original argument therefor means, that it is established that cars and robots can be safe-fied with functional safety standards and measures such that only tolerable risks exist, so they cannot be inherently unsafe, so generally they can be let be on public roads or around people with conditions.