3 ms·

I've got to retract some shade I've thrown at Waymo when discussing Tesla FSD in the past. Like others, my biggest objection with it was their approach to scal

by nunez 6mo ago

I've got to retract some shade I've thrown at Waymo when discussing Tesla FSD in the past. Like others, my biggest objection with it was their approach to scaling. Tesla aimed for the vision-only general solution with FSD, while Waymo held strong to its LIDAR-first geofence strategy. I held this opinion before using Waymo, as it wasn't available in my city (Houston, TX). I've used it several times after being invited into their Early Access earlier this month...and, wow, I couldn't have been more wrong. Waymo drives incredibly well. Like, INCREDIBLY well. Tesla FSD v14 drives well too, but Waymo feels more confident in edge case situations (of which there are many in the city driving space) and, well, I can be on my laptop or whatever during the trip. Ironically, Waymo pushed me towards using public transit in Houston, so it's incredibly sad to read that this expansion is happening as Portland's public transit system is getting defunded. The time and mental sanity I've gotten back from not driving has been immense and undeniable. (It's weird how "bus-pilled" I became after my first few Waymo trips given that I grew up in NYC taking the bus and subway all of the time.) All that said, based on how slowly Tesla is scaling their (inexcusably much more nascent) Robotaxi offering, I don't think ANY of our cars are going to get "unsupervised" FSD with the hardware they were shipped with.

TheAlchemist 6mo ago
It's good to hear that more people are realizing that Tesla is nowhere near where they claim to be. For anobody observing it from the sidelines, it was obvious for the past 3 years at least, that Tesla will not achieve real unsupervised FSD with HW3, and now it's also obvious it will not be wiht HW4. It's also obvious they are very well aware of those facts, despite lying to investors and customers for the past 10 years.
- nunez 6mo ago
  I stanned for them so hard. Oh well; live and learn. It's still the best autonomous driving system that you can buy, though now that Hyundai onlined their Waymo production line, needing a car might be a thing of the past in a few years. Maybe that's why they pivoted so hard into Optimus (at the cost of their auto division).
  - xnx 6mo ago
    It would be a great thing if Tesla somehow pulled it off, but -short of a miracle breakthrough in AGI- it is not likely.
  - decimalenough 6mo ago
    Emphasis on "can buy", because Chinese competitors like BYD's "God's Eye" (I presume it sounds better in the original Chinese) aren't available outside China. https://www.drive.com.au/news/byd-launches-driver-assistance-tech-with-ai-integration-including-for-its-cheapest-models/ https://www.drive.com.au/news/byd-launches-driver-assistance... Can't vouch for its quality since it's notoriously difficult to get reliable safety data on Chinese self-driving, but I'm reasonably sure these are the first widely available consumer vehicles with LIDAR, which is already a massive improvement on Tesla's myopic camera-only insistence.
xnx 6mo ago
This type of mea culpa is exceedingly rare. Credit to you for updating your opinion when presented with evidence, and publically admitting so.
- nunez 6mo ago
  Thanks! It's quite unfortunate that Tesla was, indeed, yanking our chains this whole time. I've been on the FSD train since I bought our first one in 2020 and have been rooting hard for them. If they were serious about being close to cracking FSD, they would double down on their auto business. Them killing their flagships (Model S/X) and investing their time on Optimus instead strongly suggests that they probably won't make it. (They did finally online their Semi production line, but I have my doubts on how well that will do.)
insane_dreamer 6mo ago
I've tried Tesla FSD a couple of times (free trial in our Model 3) and didn't like it (even though I'm an early adopter who is used to putting up with rough edges; my wife tried it once and wouldn't touch it again). It acted like a nervous new driver. Portland is a tough place to drive, lots of edge cases, so a good test case for Waymo. Having said that, as good as Waymo might be, I'm still concerned about the effect on public transport, which I use when I can. We're in the suburbs but it's good for getting downtown (no parking to worry about) and to the airport, which unlike many US cities is served by rail. We're don't live near a light rail station (wish there were more of them!) but we park and ride. I wish more people would do that, but Americans are obsessed with single car transport and the cities are designed for that.

by nunez 6mo ago

TheAlchemist 6mo ago

It's good to hear that more people are realizing that Tesla is nowhere near where they claim to be. For anobody observing it from the sidelines, it was obvious for the past 3 years at least, that Tesla will not achieve real unsupervised FSD with HW3, and now it's also obvious it will not be wiht HW4. It's also obvious they are very well aware of those facts, despite lying to investors and customers for the past 10 years.

nunez 6mo ago
I stanned for them so hard. Oh well; live and learn. It's still the best autonomous driving system that you can buy, though now that Hyundai onlined their Waymo production line, needing a car might be a thing of the past in a few years. Maybe that's why they pivoted so hard into Optimus (at the cost of their auto division).
- xnx 6mo ago
  It would be a great thing if Tesla somehow pulled it off, but -short of a miracle breakthrough in AGI- it is not likely.
- decimalenough 6mo ago
  Emphasis on "can buy", because Chinese competitors like BYD's "God's Eye" (I presume it sounds better in the original Chinese) aren't available outside China. https://www.drive.com.au/news/byd-launches-driver-assistance-tech-with-ai-integration-including-for-its-cheapest-models/ https://www.drive.com.au/news/byd-launches-driver-assistance... Can't vouch for its quality since it's notoriously difficult to get reliable safety data on Chinese self-driving, but I'm reasonably sure these are the first widely available consumer vehicles with LIDAR, which is already a massive improvement on Tesla's myopic camera-only insistence.

xnx 6mo ago

This type of mea culpa is exceedingly rare. Credit to you for updating your opinion when presented with evidence, and publically admitting so.

nunez 6mo ago
Thanks! It's quite unfortunate that Tesla was, indeed, yanking our chains this whole time. I've been on the FSD train since I bought our first one in 2020 and have been rooting hard for them. If they were serious about being close to cracking FSD, they would double down on their auto business. Them killing their flagships (Model S/X) and investing their time on Optimus instead strongly suggests that they probably won't make it. (They did finally online their Semi production line, but I have my doubts on how well that will do.)

insane_dreamer 6mo ago

I've tried Tesla FSD a couple of times (free trial in our Model 3) and didn't like it (even though I'm an early adopter who is used to putting up with rough edges; my wife tried it once and wouldn't touch it again). It acted like a nervous new driver. Portland is a tough place to drive, lots of edge cases, so a good test case for Waymo. Having said that, as good as Waymo might be, I'm still concerned about the effect on public transport, which I use when I can. We're in the suburbs but it's good for getting downtown (no parking to worry about) and to the airport, which unlike many US cities is served by rail. We're don't live near a light rail station (wish there were more of them!) but we park and ride. I wish more people would do that, but Americans are obsessed with single car transport and the cities are designed for that.