6 ms·

Just an observation .... locating the 20th century car industry in Detroit gave them an unfair advantage over California. Cars have to function correctly in ba

Just an observation .... locating the 20th century car industry in Detroit gave them an unfair advantage over California. Cars have to function correctly in bad weather -- rain, ice, snow, etc. Taxicabs especially. Imagine a transport system in a city like New York or DC where taxis don't function in a rainstorm. It's in bad weather that drivers earn their pay. I know from experience that the Tesla driver assistance stuff doesn't work in the rain. It won't engage. "Can't see well enough" is what it says. Maybe these Sili Valley types should stop fighting over fair-weather car control stuff, and spend the same money setting up development shops somewhere besides endless-summer California. Patents only last 20 years. It's going to take 30 to convert the transport network.

yawaworht12 10y ago
Like Pittsburgh?
- vermontdevil 10y ago
  Especially Pittsburgh with bridges and hills.
  - dllthomas 10y ago
    I think the point was that Uber's self-driving effort is located in Pittsburgh, so when the parent says > Maybe these Sili Valley types should stop fighting over fair-weather car control stuff, and spend the same money setting up development shops somewhere besides endless-summer California. that's exactly what Uber's done.
- tsunamifury 10y ago
  We get the it, you work for Cruise.
- djsumdog 10y ago
  Where an Uber autonomous vehicle went the wrong way up a 1-way street?
Fricken 10y ago
Waymo has been testing up around Lake Tahoe in the snow, and in Arizona to deal with dust, and in Washington to deal with rain. Ford and GM are both testing in Michigan. Here's Drive.ai demoing L4 in a downpour at night (really impressive): https://youtu.be/GMvgtPN2IBU https://youtu.be/GMvgtPN2IBU But in general terms it's logical to get everything working well in optimal conditions before moving on to challenging weather. I don't think they consider any of this stuff to be unsolvable, short of maybe a whiteout blizzard, but, you know, first things first.
lern_too_spel 10y ago
Nobody has the capital to set up a taxi system across the US simultaneously. Even Uber, which doesn't have to pay for the cars up front, has to launch a city at a time. Self driving taxis will follow a similar path, launching in Mountain View by the end of the year (according to Alphabet's original claims), expanding to Silicon Valley and LA, and then eventually expanding everywhere. Owning personal transportation in Silicon Valley alone is a nearly billion dollar opportunity.
bsder 10y ago
> Just an observation .... locating the 20th century car industry in Detroit gave them an unfair advantage over California. Cars have to function correctly in bad weather -- rain, ice, snow, etc. Um, that's fanciful thinking. Especially since early cars were notoriously uncomfortable and unreliable. Detroit had three things going for it. The first was proximity. Remember, this is 1900. Transport by rail and ship is primary. Detroit is directly on both between Chicago and New York. And the entire Northeast simply dwarfs the rest of the country in population at the time: http://etc.usf.edu/maps/pages/2000/2064/2064.htm http://etc.usf.edu/maps/pages/2000/2064/2064.htm The second was Ransom Eli Olds. Detroit was the Silicon Valley of its day because the equivalent of Fairchild/The Traitorous Eight all spawned out from companies that were associated with Oldsmobile or Olds, himself.
bsder 10y ago
> It's going to take 30 to convert the transport network. Not even close. AI will be better than humans at driving almost immediately. Here's a little inside tip: Humans SUCK at driving. It is a boring, monotonous task that demands full attention suddenly and randomly. Humans make very poor decisions in these changeover situations. Bad weather, bad signage, and other bad drivers confound human drivers at every turn.
- yawaworht12 10y ago
  Completely agree. The second and more important problem is that humans suck worse at legislation and governance. I expect self driving cars to take off in developing countries and forward thinking countries (like Singapore) well before they take off in the United States.
djsumdog 10y ago
> It's going to take 30 to convert the transport network. I think the tech is amazing from a software standpoint, but the amount of resources being dumped into self-driving cars could be better spent fixing our broken (non-existent) mass transit system in America first. I wrote a post on this a few months ago: http://penguindreams.org/blog/self-driving-cars-will-not-solve-the-transportation-problem/ http://penguindreams.org/blog/self-driving-cars-will-not-sol...

yawaworht12 10y ago

Like Pittsburgh?

vermontdevil 10y ago
Especially Pittsburgh with bridges and hills.
- dllthomas 10y ago
  I think the point was that Uber's self-driving effort is located in Pittsburgh, so when the parent says > Maybe these Sili Valley types should stop fighting over fair-weather car control stuff, and spend the same money setting up development shops somewhere besides endless-summer California. that's exactly what Uber's done.
tsunamifury 10y ago
We get the it, you work for Cruise.
djsumdog 10y ago
Where an Uber autonomous vehicle went the wrong way up a 1-way street?

Fricken 10y ago

Waymo has been testing up around Lake Tahoe in the snow, and in Arizona to deal with dust, and in Washington to deal with rain. Ford and GM are both testing in Michigan. Here's Drive.ai demoing L4 in a downpour at night (really impressive): https://youtu.be/GMvgtPN2IBU https://youtu.be/GMvgtPN2IBU But in general terms it's logical to get everything working well in optimal conditions before moving on to challenging weather. I don't think they consider any of this stuff to be unsolvable, short of maybe a whiteout blizzard, but, you know, first things first.

lern_too_spel 10y ago

Nobody has the capital to set up a taxi system across the US simultaneously. Even Uber, which doesn't have to pay for the cars up front, has to launch a city at a time. Self driving taxis will follow a similar path, launching in Mountain View by the end of the year (according to Alphabet's original claims), expanding to Silicon Valley and LA, and then eventually expanding everywhere. Owning personal transportation in Silicon Valley alone is a nearly billion dollar opportunity.

bsder 10y ago

> Just an observation .... locating the 20th century car industry in Detroit gave them an unfair advantage over California. Cars have to function correctly in bad weather -- rain, ice, snow, etc. Um, that's fanciful thinking. Especially since early cars were notoriously uncomfortable and unreliable. Detroit had three things going for it. The first was proximity. Remember, this is 1900. Transport by rail and ship is primary. Detroit is directly on both between Chicago and New York. And the entire Northeast simply dwarfs the rest of the country in population at the time: http://etc.usf.edu/maps/pages/2000/2064/2064.htm http://etc.usf.edu/maps/pages/2000/2064/2064.htm The second was Ransom Eli Olds. Detroit was the Silicon Valley of its day because the equivalent of Fairchild/The Traitorous Eight all spawned out from companies that were associated with Oldsmobile or Olds, himself.

bsder 10y ago

> It's going to take 30 to convert the transport network. Not even close. AI will be better than humans at driving almost immediately. Here's a little inside tip: Humans SUCK at driving. It is a boring, monotonous task that demands full attention suddenly and randomly. Humans make very poor decisions in these changeover situations. Bad weather, bad signage, and other bad drivers confound human drivers at every turn.

yawaworht12 10y ago
Completely agree. The second and more important problem is that humans suck worse at legislation and governance. I expect self driving cars to take off in developing countries and forward thinking countries (like Singapore) well before they take off in the United States.

djsumdog 10y ago

> It's going to take 30 to convert the transport network. I think the tech is amazing from a software standpoint, but the amount of resources being dumped into self-driving cars could be better spent fixing our broken (non-existent) mass transit system in America first. I wrote a post on this a few months ago: http://penguindreams.org/blog/self-driving-cars-will-not-solve-the-transportation-problem/ http://penguindreams.org/blog/self-driving-cars-will-not-sol...