17 ms·

Just a few months ago, I recall many skeptics on HN claiming that full driverless technology was 10-20 years away. Do those people still think that it's going

Just a few months ago, I recall many skeptics on HN claiming that full driverless technology was 10-20 years away. Do those people still think that it's going to take 10-20 years for this tech to reach say, New York, now that it's live in Phoenix?

pb7 6y ago
It might actually. Phoenix suburbs are different than the densest city in the US in a region that sees the worst of weather. Saying this as someone that hopes Waymo succeeds.
chubot 6y ago
Yes, because this announcement is a lot like their previous announcements -- fudging the language to make it sound like it's more available than it is. It's not even generally available in Phoenix yet. I'm sure someone on HN lives in Phoenix, and has never used Waymo before. Tell us if you can get a driverless ride. It's got a long way to go before it's something like Uber in Phoenix. It's got even longer to go before it's something like Uber in NYC (10-20 years still sounds right). ---- The funny thing is that the pandemic should have been HUGE for self-driving cars... Most people I know haven't taken rideshare in 6+ months because they don't want to ride with somebody, and I only did so in the last few weeeks. But it was a non-event. I didn't hear anyone talking about self-driving in the last 6 months. If it really worked, and was really available, then there would be certainly some people who would pay 2-5x the price of Lyft/Uber for no driver. (Not most people, but some people.)
- pb7 6y ago
  I don't see the relationship between COVID and increased demand for self-driving. You're still sharing the same vehicle with potentially hundreds of people a day that are touching all the same surfaces. The driver is only a vector if they become infected. Other than that, the highest risk is touching doors and such.
  - vernon99 6y ago
    This is an scientific take and most of the casual folks don’t see it this way.
  - chubot 6y ago
    The car can be cleaned either way -- with a driver or without. That's what happens in grocery stores and hotels now. They wipe down the shopping carts, and they seal the doors. But there's only one way remove the possibility of getting sick from the driver. We've been talking about that for 15 years, but it doesn't work yet. I don't think the risks are equivalent on a per-person basis either. The "viral load" apparently matters, so using the same car as 1 person is not the same as riding with a sick driver for say 30 minutes. It could be that the driver risk is great than that of 30 passengers combined.
- RivieraKid 6y ago
  Will you cease to be a skeptic once Waymo One is open to anyone?
  - chubot 6y ago
    Weird question :) My claim is that it won't be open to "anyone" in a useful way any time soon. And you need to define "anyone" -- that's precisely what's being fudged here (and what's been fudged in previous announcements). To be clear: I think Waymo will fold or be wound down before this happens. In a previous comment, I said that IF we get economically feasible self-driving cars in the next 10-20 years, then it will not be from Waymo. It will be from a breakthrough out of left field. In other words, "we" will have thrown out millions of lines of code and billions of dollars of investment, and started over. Obviously, if they roll it out, then I'll be wrong! But if I thought that was possible now, then I wouldn't bother calling them out on it. Previous comments: https://news.ycombinator.com/item?id=22632588 https://news.ycombinator.com/item?id=22632588
    - twmahna 6y ago
      Interesting take! Why does this perspective make you bullish on Uber/Lyft? Isn't a large portion of their FCF tied to reducing labor costs by substituting drivers with self-driving tech? AB5 and Democrats potentially gaining unified control of Washington don't bode well for their cost structure either.
      - chubot 6y ago
        I think Uber and Lyft can raise their prices significantly and people will still take them, simply because the alternatives are worse. Owning a car is worse for many situations, and so is public transportation. If you have any friends with 16-18 year old children, ask them if they are going to buy them a car, or if they'd rather just pay for Uber/Lyft (that's an honest question, but I know what my experience is.) Ditto for people taking care of senior citizens. There is a fundamental efficiency to rideshare. And it will continue to get better. As I said in the prior comment -- we already have self-driving cars. I push a button on my phone and somebody drives me somewhere. For now, the self-driving tech is just a big distraction that makes things fantastically more expensive and unreliable. (This could change, but again I see that as 20-30 years in the future. I don't price it in for sure) Also, drivers seem to love Uber and Lyft for the work flexibility, at least when I talked to many of them in 2015-2018. Those jobs were really valuable to a diverse set of people. ----- I've heard the viewpoints you quoted in the media, but I don't think they are painting a balanced picture. It seems like a message that somebody wants to get out there, not a neutral viewpoint. Of course the whole idea that Uber would reduce their costs with self-driving was also a fantasy meant to sell stock. I don't think the market ever priced that in. It's clear now that this area was a disaster, and the stock didn't crash because of that. I think Uber/Lyft will be creating value for 10-20 years to come. People in SF seem to be bearish on them, but I think that's sort of provincial view. We might be on to the next big thing, but valuable tech is rare, and it takes years/decades for valuable technology to be adopted everywhere that it makes sense. People are still buying iPhones for the first time, even though it was released in 2007. It's hard for me to imagine, but there must be people using Google for the first time too in 2020.
air7 6y ago
Imo not even in 20. Driving occasionally requires an AGI level understanding of humans and physics which is out of reach currently.
YeGoblynQueenne 6y ago
Well, I don't know who "these people" are and what exactly they thought, but I don't see why they wouldn't still be skeptical. The announcement promises that "in the near term" all Waymo's rides will be "fully driverless" but in the immediate term they are modifying their cars to go back to rides with a safety driver. So the "near term" will be fully driverless, but right now it's even less drivereless then before. For myself, I would trust this announcement a lot more if it claimed some new technological breakthrough that enables Waymo to do now something that wasn't possible half a year, or 5 or 10 years ago (specifically, this "something" is level-3 self-driving, as poorly as this is unfortunately defined). I don't see any such claims. The announcement instead reads as so much marketing copy to me, and it is not saying anything that has any practical implications for the technological capabilities of Waymo's cars.

pb7 6y ago

It might actually. Phoenix suburbs are different than the densest city in the US in a region that sees the worst of weather. Saying this as someone that hopes Waymo succeeds.

chubot 6y ago

Yes, because this announcement is a lot like their previous announcements -- fudging the language to make it sound like it's more available than it is. It's not even generally available in Phoenix yet. I'm sure someone on HN lives in Phoenix, and has never used Waymo before. Tell us if you can get a driverless ride. It's got a long way to go before it's something like Uber in Phoenix. It's got even longer to go before it's something like Uber in NYC (10-20 years still sounds right). ---- The funny thing is that the pandemic should have been HUGE for self-driving cars... Most people I know haven't taken rideshare in 6+ months because they don't want to ride with somebody, and I only did so in the last few weeeks. But it was a non-event. I didn't hear anyone talking about self-driving in the last 6 months. If it really worked, and was really available, then there would be certainly some people who would pay 2-5x the price of Lyft/Uber for no driver. (Not most people, but some people.)

pb7 6y ago
I don't see the relationship between COVID and increased demand for self-driving. You're still sharing the same vehicle with potentially hundreds of people a day that are touching all the same surfaces. The driver is only a vector if they become infected. Other than that, the highest risk is touching doors and such.
- vernon99 6y ago
  This is an scientific take and most of the casual folks don’t see it this way.
- chubot 6y ago
  The car can be cleaned either way -- with a driver or without. That's what happens in grocery stores and hotels now. They wipe down the shopping carts, and they seal the doors. But there's only one way remove the possibility of getting sick from the driver. We've been talking about that for 15 years, but it doesn't work yet. I don't think the risks are equivalent on a per-person basis either. The "viral load" apparently matters, so using the same car as 1 person is not the same as riding with a sick driver for say 30 minutes. It could be that the driver risk is great than that of 30 passengers combined.
RivieraKid 6y ago
Will you cease to be a skeptic once Waymo One is open to anyone?
- chubot 6y ago
  Weird question :) My claim is that it won't be open to "anyone" in a useful way any time soon. And you need to define "anyone" -- that's precisely what's being fudged here (and what's been fudged in previous announcements). To be clear: I think Waymo will fold or be wound down before this happens. In a previous comment, I said that IF we get economically feasible self-driving cars in the next 10-20 years, then it will not be from Waymo. It will be from a breakthrough out of left field. In other words, "we" will have thrown out millions of lines of code and billions of dollars of investment, and started over. Obviously, if they roll it out, then I'll be wrong! But if I thought that was possible now, then I wouldn't bother calling them out on it. Previous comments: https://news.ycombinator.com/item?id=22632588 https://news.ycombinator.com/item?id=22632588
  - twmahna 6y ago
    Interesting take! Why does this perspective make you bullish on Uber/Lyft? Isn't a large portion of their FCF tied to reducing labor costs by substituting drivers with self-driving tech? AB5 and Democrats potentially gaining unified control of Washington don't bode well for their cost structure either.
    - chubot 6y ago
      I think Uber and Lyft can raise their prices significantly and people will still take them, simply because the alternatives are worse. Owning a car is worse for many situations, and so is public transportation. If you have any friends with 16-18 year old children, ask them if they are going to buy them a car, or if they'd rather just pay for Uber/Lyft (that's an honest question, but I know what my experience is.) Ditto for people taking care of senior citizens. There is a fundamental efficiency to rideshare. And it will continue to get better. As I said in the prior comment -- we already have self-driving cars. I push a button on my phone and somebody drives me somewhere. For now, the self-driving tech is just a big distraction that makes things fantastically more expensive and unreliable. (This could change, but again I see that as 20-30 years in the future. I don't price it in for sure) Also, drivers seem to love Uber and Lyft for the work flexibility, at least when I talked to many of them in 2015-2018. Those jobs were really valuable to a diverse set of people. ----- I've heard the viewpoints you quoted in the media, but I don't think they are painting a balanced picture. It seems like a message that somebody wants to get out there, not a neutral viewpoint. Of course the whole idea that Uber would reduce their costs with self-driving was also a fantasy meant to sell stock. I don't think the market ever priced that in. It's clear now that this area was a disaster, and the stock didn't crash because of that. I think Uber/Lyft will be creating value for 10-20 years to come. People in SF seem to be bearish on them, but I think that's sort of provincial view. We might be on to the next big thing, but valuable tech is rare, and it takes years/decades for valuable technology to be adopted everywhere that it makes sense. People are still buying iPhones for the first time, even though it was released in 2007. It's hard for me to imagine, but there must be people using Google for the first time too in 2020.

air7 6y ago

Imo not even in 20. Driving occasionally requires an AGI level understanding of humans and physics which is out of reach currently.

YeGoblynQueenne 6y ago

Well, I don't know who "these people" are and what exactly they thought, but I don't see why they wouldn't still be skeptical. The announcement promises that "in the near term" all Waymo's rides will be "fully driverless" but in the immediate term they are modifying their cars to go back to rides with a safety driver. So the "near term" will be fully driverless, but right now it's even less drivereless then before. For myself, I would trust this announcement a lot more if it claimed some new technological breakthrough that enables Waymo to do now something that wasn't possible half a year, or 5 or 10 years ago (specifically, this "something" is level-3 self-driving, as poorly as this is unfortunately defined). I don't see any such claims. The announcement instead reads as so much marketing copy to me, and it is not saying anything that has any practical implications for the technological capabilities of Waymo's cars.