4 ms·

Hmm, on a related thought, how could self driving cars ever work there?

naravara 10y ago
They won't. Firstly because the traffic is insane. Secondly because labor rates are so low that rickshaw and taxi drivers are dirt cheap. The productivity function is way skewed towards labor in India to where expensive robots to drive the cars don't make economic sense.
- xapata 10y ago
  Won't in the short run, perhaps. Sure, in the long run we're all dead, but somewhere in the middle there will be robots and automobiles in the more complete sense of the word.
- xapata 10y ago
  Won't in the short run, perhaps. Sure, in the long run we're all dead, but somewhere in the middle there will be robots and auto-mobiles in the more Latin sense of the word.
joshjje 10y ago
I would think they would work even better there?
- djsumdog 10y ago
  Highly doubtful. There are way too many variables. If they worked at all, they'd be unable to move or only at a speed of <5kph. It's simply too chaotic and way too many variables.
  - joshjje 9y ago
    I see what you are saying, but I still think a computer would do better in an environment like that with less rules than say the US. At least with the goal being not causing accidents. A computer with 360 vision is undoubtedly better at avoiding accidents probably in any condition. It may result in as you say too slow speeds or paralyzation, but many "rules" in the US either arent written down, or often arent followed making that situation more likely than with less rules imo.
hsuresh 10y ago
If it follows other tech trends(eg: cell phones bypassing land lines), India and other developing countries will see self-driving cars before freeways and owned cars. Well, I'm an optimist.
- naravara 10y ago
  Maybe. But I'm not sure it's as easy as putting things in a scale from more modern to less modern. The places where India has leapfrogged has been on stuff that needs new infrastructure development since there was never any good old infrastructure to take up resources or incumbent players to choke out new players. Self driving cars don't need new infrastructure, its functionally an application that runs on existing infrastructure. If you have decent roads it'll work the same way whether you're in India or anywhere else. Although I guess you could argue that Indians are less prissy about what they're willing to tolerate. A self-driving auto-rickshaw wouldn't fly in America, but it would be just fine for just about any Indian, and it's a much more useful application of self-driving tech than expecting everyone to have their own robo-chauffeured luxury cars.
- toast0 10y ago
  Cell phones bypassing land lines makes sense because you can cover more customers with less cost. Self-driving cars may eventually be less costly than human driven cars, but for the near future, the costs you add (sensors, software, actuators, etc) are way more expensive than the costs you save (driver's attention, optimistically damage to persons and property reduced through collision avoidance), particularly in developing countries where a person's attention has less monetary value, but the raw materials for the car have similar costs. It's unlikely that you'd get much of the ancillary benefits of usability by unlicensed drivers either -- few people will be able to afford a driverless car to cart their kid around to school or wherever, anyway those people could just hire a driver and a normal car.

They won't. Firstly because the traffic is insane. Secondly because labor rates are so low that rickshaw and taxi drivers are dirt cheap. The productivity function is way skewed towards labor in India to where expensive robots to drive the cars don't make economic sense.

xapata 10y ago
Won't in the short run, perhaps. Sure, in the long run we're all dead, but somewhere in the middle there will be robots and automobiles in the more complete sense of the word.
xapata 10y ago
Won't in the short run, perhaps. Sure, in the long run we're all dead, but somewhere in the middle there will be robots and auto-mobiles in the more Latin sense of the word.

joshjje 10y ago

I would think they would work even better there?

djsumdog 10y ago
Highly doubtful. There are way too many variables. If they worked at all, they'd be unable to move or only at a speed of <5kph. It's simply too chaotic and way too many variables.
- joshjje 9y ago
  I see what you are saying, but I still think a computer would do better in an environment like that with less rules than say the US. At least with the goal being not causing accidents. A computer with 360 vision is undoubtedly better at avoiding accidents probably in any condition. It may result in as you say too slow speeds or paralyzation, but many "rules" in the US either arent written down, or often arent followed making that situation more likely than with less rules imo.

hsuresh 10y ago

If it follows other tech trends(eg: cell phones bypassing land lines), India and other developing countries will see self-driving cars before freeways and owned cars. Well, I'm an optimist.

naravara 10y ago
Maybe. But I'm not sure it's as easy as putting things in a scale from more modern to less modern. The places where India has leapfrogged has been on stuff that needs new infrastructure development since there was never any good old infrastructure to take up resources or incumbent players to choke out new players. Self driving cars don't need new infrastructure, its functionally an application that runs on existing infrastructure. If you have decent roads it'll work the same way whether you're in India or anywhere else. Although I guess you could argue that Indians are less prissy about what they're willing to tolerate. A self-driving auto-rickshaw wouldn't fly in America, but it would be just fine for just about any Indian, and it's a much more useful application of self-driving tech than expecting everyone to have their own robo-chauffeured luxury cars.
toast0 10y ago
Cell phones bypassing land lines makes sense because you can cover more customers with less cost. Self-driving cars may eventually be less costly than human driven cars, but for the near future, the costs you add (sensors, software, actuators, etc) are way more expensive than the costs you save (driver's attention, optimistically damage to persons and property reduced through collision avoidance), particularly in developing countries where a person's attention has less monetary value, but the raw materials for the car have similar costs. It's unlikely that you'd get much of the ancillary benefits of usability by unlicensed drivers either -- few people will be able to afford a driverless car to cart their kid around to school or wherever, anyway those people could just hire a driver and a normal car.