Why is WAYMO so much better than FSD ?

[cyberos]

Honestly? Like better means I watch a car next to me drive by me with no one in it in a collage campus area and 17 kids crossing a street after the light is red while I hav to correct my truck with me inside it ? Define better ? .

Like I said, people either accepted or rejected LIDAR as the answer to your question

Tesla FSD can hop curbs and terrorize roundabouts in every corner of America and Waymo can only drive flawlessly in four cities

The consensus is that Tesla is “winning” because of machine learning from the constantly growing supply of video examples from drivers like us. Waymo’s rate of training is orders of magnitude slower and complicated by multiple data formats (video+LIDAR+others)

 

[nico007]

I don’t know who is winning between Waymo and Tesla but this weekend I have been doing Oakland to Monterey and return door to door with 0 intervention, 0 scare, perfect speed (when the road is empty and straight the car sets at 79) and very smooth ride.
FSD user since day 1 and it’s never been that good.

 

[YDR37]

The consensus is that Tesla is “winning” because of machine learning from the constantly growing supply of video examples from drivers like us. Waymo’s rate of training is orders of magnitude slower and complicated by multiple data formats (video+LIDAR+others)

For Waymo to "win", they would have add a lot more service areas. The bad news (for Waymo) is that such expansion is expensive and is going slowly. The good news (for Waymo) is that there are no apparent technological barriers to such expansion. Mapping out a new service area undoubtedly costs a lot of money and takes a lot of work, but it's clearly do-able if Waymo throws enough people and cash at the problem. So Waymo's path to victory appears straightforward, but it also looks like a slow and expensive slog.

For Tesla to "win", FSD would have to get a lot more reliable. The good news (for Tesla) is that FSD has improved rapidly over the past few years. The bad news (for Tesla) is that there is no guarantee that FSD will ever improve enough to meet reliability goals. Tesla is undoubtedly throwing lots of people and cash at the problem, and maybe the effort will lead to a breakthrough-- or maybe the effort will crash into a technological barrier with diminishing returns. So Tesla's path to victory involves more risk: in the best case, it could be much shorter than Waymo's; in the worst case, it could be a dead end.

 

[AceMcLoud]

Waymo is not better because its strictly geo-fenced to only operate in known areas. It also often locks up and needs help from a remote operator in really easy situations. It's what you would expect from a system that is hard coded. They took a short-cut path to autonomy that is not able to scale to the rest of the world without spending billions on further mapping and training for each specific area they want to operate in. It's a "dumb" system. It can't even go on the Interstate, it just bumbles along on surface streets.

Waymo doesn't miss streets because they are all hard-coded. At least in the few areas Waymo has hard coded and trained on. It's definitely not the future of autonomy in the way that FSD is. Neither are ready for prime time, and Waymo never will unless they adopt a more natural way of driving.

Yeah, yeah, the magical Waymo maps that include current traffic. You know nothing about the system yet think you're ready to have an opinion.

Trump voter?