Doggydogworld
Active Member
Because it will be good enough in April 2020 that you can sleep while it drives. *Checks date*..... oh, oops.
-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Autonomous Car Progress
- Thread starter stopcrazypp
- Start date
-
- Tags
- Autonomous Vehicles
EVNow
Well-Known Member
That’s obsolete … you need to keep up with newer tweets
Afterall, according to Waymo/Google nobody would take a driving test now because robotaxis would be so ubiquitous.
Daniel in SD
(supervised)
No one is claiming that Waymo runs in your neighborhood but you mention it all the time as a shortcoming! Geographical coverage is a fair metric for Waymo, though I would say it's not nearly as important as achieving safe driverless operation at low enough cost.Because *no one* is claiming it is an unsupervised system *now*.
Its like trying to use the current SpaceX rocket to fly to Mars .... after all that's the objective, right ?
Afterall, Waymo won't run in my neighborhood ...
EVNow
Well-Known Member
You are an expert at moving goal posts.
Going back to my original point, it makes zero sense to say Early Access FSD Beta is not at human level. Duh.
Daniel in SD
(supervised)
This makes zero sense, no one is claiming that I'm not an expert at moving goal posts!
Anyway the statement was that FSD Beta isn't safe enough anywhere which is different than saying it's not at human level. One would expect it would achieve human performance some places before others while it's still in beta.
EVNow
Well-Known Member
I’m. Are you saying I’m “no one”?This makes zero sense, no one is claiming that I'm not an expert at moving goal posts!
No, I would not expect it to be at disengagement rates of 1 in 10 miles in some places and 1 in 10,000 in some others. That is inherently an idea of geo-fencers.
ps :
diplomat33
Average guy who loves autonomous vehicles
Daniel in SD
(supervised)
No one is claiming that I'm not an expert at moving goalposts therefore there was no reason for you state that I am by your logic.I’m. Are you saying I’m “no one”?This makes zero sense, no one is claiming that I'm not an expert at moving goal posts!
I agree. Of course that is the argument people use against other SDC companies. They can get 1 in 10,000 miles per disengagement in their geofenced area but if they were to test somewhere else it would be worse than Tesla.
This is a Tesla forum so people are always going to be comparing other company's self-driving systems to FSD.
bradtem
Robocar consultant
Which teams are you referring to. Do you have data to back this up? Are you making a strange strawman argument by suggesting they would not map the new area first? MobilEye sent 2 employees to Munich for 2 weeks to make sure their cars were working fine there and to polish the maps, and were up and running at a rate vastly better than Tesla, as far as I understand. I am sure that when Waymo moves to new areas they are vastly, vastly better than Tesla too. So what do you base your view on? Not using maps is a bug that makes you drive poorly everywhere, not a feature.
EVNow
Well-Known Member
No - if they try it in my neighborhood it would not work at all, since they (probably) don't have HD Maps
Proves my point - that you are adept at moving goal posts.
EVNow
Well-Known Member
Daniel in SD
(supervised)
No, sorry for all the silly bickering on the thread.
I was saying that it's an argument that I've seen other people make. I think it's absurd.
diplomat33
Average guy who loves autonomous vehicles
This is a Tesla misconception based on the false premise that Waymo is only doing specialized FSD, ie geohacking FSD to only work for a specific area. But this is false. The stack that Waymo uses is generalized. For example, it would not be worse at detecting objects or predicting objects in different areas. In fact, I would expect Waymo would be better than Tesla's FSD Beta in areas that are not premapped because of Waymo's sophisticated perception and planning, but not good enough for driverless (hence why Waymo always premaps). Of course, there could be edge cases in the new area that could cause the Waymo Driver to perform slightly worse but definitely not worse than Tesla. I would expect the Waymo Driver to be fairly consistently good everywhere that is premapped, but again probably not good enough for driverless (why Waymo is not scaling yet, since "fairly good everywhere" is not good enough for driverless robotaxis).
Last edited:
That's not what they said. He specifically mentioned his son and i'm sure his son would live with him and that he lives where their development team hence would have access to the cars.That’s obsolete … you need to keep up with newer tweets
Afterall, according to Waymo/Google nobody would take a driving test now because robotaxis would be so ubiquitous.
EVNow
Well-Known Member
LOL. Yes, he was specifically talked about this one kid and it would not apply to anyone else and it didn’t mean anything other than that one rich kid wouldn’t need to take a driving test.
Very believable.
EVNow
Well-Known Member
diplomat33
Average guy who loves autonomous vehicles
On Instagram, Mobileye posted a timeline that shows their biggest milestones. Here is the timeline for the last few years that has some of their most recent FSD milestones. I thought it was informative in case anyone was wondering what Mobileye has been doing in FSD lately.
EVNow
Well-Known Member
Within few minutes of driving it starts beeping (or may be it is edited) - see 8:21. Not sure why ....
ps : Apparently the car in front is "too far away" ? Got to love Level 3.
Last edited:
Not to ruin Master Plan 3 for everybody, but the future of FSD won't need cameras or processors in your car. Instead a Tesla AI robot will get into the car and drive you through Boring Company tunnels to where you want to go, all tracked by SpaceX Starlink satellites. Of course, the new cost of FSD will go to $50,000.
Similar threads
- Article
