It's less of an adoption lag problem and more of a ubiquity and affordability problem vs vehicle lifespans.
Average price of a new car with a vision system is ~$80k, 90% of the population can't afford that. Tesla is great and more affordable, but they've made ~7m total cars and there are 1.5bn on the road around the world with 350m on US roads. What's the total number of autonomous vehicles on US roads between Waymo, Zoox, and all the others? ~5000? Even if it was 500,000; that's a drop in the 350m car bucket.
The national fleet takes 30 years to cycle, so even if every car produced today had a vision system (it's ~20% of new production tops) we won't have ubiquity until 2056.
The only path to eliminating all road deaths is a vision system in every car. That's what we're doing at Saferide.ai
I think a large portion of the lagging adoption is that the people that drive for Uber or Lyft tend to be illegal immigrants on fake identifications of the apps since that role is much easier to get than one that requires real documentation and an actual w2.
Didn't really read the article, but what I tend to find very hard to motivate is the safety aspect of self driving. I just think it's not very compelling. Considering the insane amount of investments necessary to even get to a mostly autonomous country-wide car fleet and the "few" people who die each year in traffic, I just think those dollars are put elsewhere much more efficiently. I don't have anything against autonomous driving, I think it's really cool. I just find the safety aspect of it super uncompelling
Yeah, but my point is that if you want to be a Peter Singer then it‘s best you start with less money in a third world country because there it‘ll be much more cost effective to save lives, if this is what you‘re after
Since Waymo and the like drive significantly less than human drivers are you making the claim that the safety numbers will remain the same when robo cars scale? Or are you choosing to not address the issue of using safety numbers from a small and controlled dataset and extrapolating that to all driving nationwide.
It's less of an adoption lag problem and more of a ubiquity and affordability problem vs vehicle lifespans.
Average price of a new car with a vision system is ~$80k, 90% of the population can't afford that. Tesla is great and more affordable, but they've made ~7m total cars and there are 1.5bn on the road around the world with 350m on US roads. What's the total number of autonomous vehicles on US roads between Waymo, Zoox, and all the others? ~5000? Even if it was 500,000; that's a drop in the 350m car bucket.
The national fleet takes 30 years to cycle, so even if every car produced today had a vision system (it's ~20% of new production tops) we won't have ubiquity until 2056.
The only path to eliminating all road deaths is a vision system in every car. That's what we're doing at Saferide.ai
I think a large portion of the lagging adoption is that the people that drive for Uber or Lyft tend to be illegal immigrants on fake identifications of the apps since that role is much easier to get than one that requires real documentation and an actual w2.
Didn't really read the article, but what I tend to find very hard to motivate is the safety aspect of self driving. I just think it's not very compelling. Considering the insane amount of investments necessary to even get to a mostly autonomous country-wide car fleet and the "few" people who die each year in traffic, I just think those dollars are put elsewhere much more efficiently. I don't have anything against autonomous driving, I think it's really cool. I just find the safety aspect of it super uncompelling
The current Value if a Statistical Life is 13.7 million.
https://www.transportation.gov/office-policy/transportation-policy/revised-departmental-guidance-on-valuation-of-a-statistical-life-in-economic-analysis
40,000 fatalities in the US is a 550 billion cost to society, annually. That does not include the cost of serious injury.
The safety aspect is real. Especially to the families of those ‘few’ that you refer to.
There is another unseen factor in the current road safety infrastructure doesn’t work very well for the growing weight of vehicles.
https://youtu.be/ZLwMroMmpC4?si=_exo1lABQAE5gFnd
The cost to upgrade the road safety infrastructure to a level to accommodate the changing vehicle fleet is in the trillions.
Yeah, but my point is that if you want to be a Peter Singer then it‘s best you start with less money in a third world country because there it‘ll be much more cost effective to save lives, if this is what you‘re after
Since Waymo and the like drive significantly less than human drivers are you making the claim that the safety numbers will remain the same when robo cars scale? Or are you choosing to not address the issue of using safety numbers from a small and controlled dataset and extrapolating that to all driving nationwide.