Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register
This site may earn commission on affiliate links.
My first year of ownership of my Model 3, I was obsessed with seeing how efficient I could get my Wh/mi numbers. I remember setting follow distance to 1 and then finding a tractor trailer to follow that was going my typical highway speed (around 70mph). I would get about 30Wh/mi improvement drafting off the semi.

I know all about this phenomenon biking with my triathlon club, but when you're in a car, you don't really feel the efficiency gain because your legs don't get tired from moving the car.
This is partly why I want autonomous cars.

If they could all form a train it would be so much more efficient. The train of vehicles would automatically break apart when one of them needed to take an exit, and then they'd rejoin.
 
What I find funny about that list is it says "Autosteer on City Streets"

But, we all know FSD Beta is more than Autosteer on City Streets.

Now it's missing a LOT of components to even be a beta version of Robotaxi. But, at the same time its more than Autosteer on City Streets.

Whoever is in charge of the website needs to sit down with Elon, and the AP/FSD Engineers to get on the same page as to what they're doing.

It is possible that FSD Beta is the testing of the autonomous stack but that the public release of "autosteer on city streets" will be a stripped down version of FSD Beta. So, Tesla could release the parts of FSD Beta, as "autosteer on city streets", that they feel are reliable enough for full public release. It would allow them to say that they fulfilled the promise of releasing "autosteer on city streets" while they are still working to make FSD Beta reliable enough.

I can kind of see why Tesla calls it "autosteer on city streets". I think the idea is that "autosteer" does more than just steer to stay in the lane, "autosteer" can now turn the wheel all the way to 90 degrees to make turns at intersections. Additionally, "autosteer" is capable of steering to stay in the lane on residential streets with no lane lines or steer around parked cars.

Personally, I think of FSD Beta as "City NOA" as the main feature seems to be that it tries to navigate a route to get you to a destination that you put in. It does auto lane changes to stay on route, auto lane changes to pass slow traffic and makes turns at intersections to follow a route which is similar to what NOA does, except it does it for city streets instead of highways.

Interestingly, "Autosteer on city streets" used to be called "City NOA" back in 2019 when Tesla added a list of features to the FSD page. Personally, I feel like "autosteer on city streets" sounds less autonomous than "City NOA". It seems that Tesla might have changed the name to downplay the autonomy of the feature, perhaps to avoid criticism when the feature is released and it falls short of L4/L5 autonomy.
 
Are you sure? It was definitely "Automatic Driving on City Streets" for awhile.

You are correct. I made a mistake. I was remembering it wrong. "City NOA" was the informal name that some on this forum used at the time. It used to be "automatic driving on city streets" and then they changed it to "autosteer on city streets" which IMO sounds less autonomous than "automatic driving on city streets.

Thanks for the correction.
 
This is partly why I want autonomous cars. If they could all form a train it would be so much more efficient.
I was upset by the comment against tailgating too, glad you commented.

I am kinda shocked that Elon mentioned during the Semi announcement that they could do this already, but we have seen no progress in it. Perhaps it would be another thing to get by regulators, and perhaps they figure Tesla drivers are too pig headed (we are) to wanna go at someone else's pace, but driving down Hwy 5 all the trucks are doing it, and we have the long East Bay and Peninsula commutes near me that could easily coordinate a collection of these trains.

I would LOVE to be able to tell my Tesla to make a train behind my RV since I cannot tow it w/o putting it up on a trailer.

Or better yet, how about people ride in ONE car and their cars all follow like a train, so they can have conversation, and when they need to go separate ways they stop and do a Chinese Fire Drill (wait, is that racist? Boy I was raised wrong) and you are on your way.
 
FSD working well enough to launch a ubiquitous Robotaxi service seems seriously so far away, I can't even picture it right now.

Heavily geofenced to certain areas/roads and weather conditions and not requiring seamless parking before picking up passengers and dropping them off, that could maybe happen in the foreseeable future if Tesla is already working on the booking/payment functions and any regulatory hurdles.

But even just geofencing across the whole US would be such a monumental task that it doesn't seem realistic, they're too far down the path of generalized autonomy
 
FSD working well enough to launch a ubiquitous Robotaxi service seems seriously so far away, I can't even picture it right now.

Heavily geofenced to certain areas/roads and weather conditions and not requiring seamless parking before picking up passengers and dropping them off, that could maybe happen in the foreseeable future if Tesla is already working on the booking/payment functions and any regulatory hurdles.

But even just geofencing across the whole US would be such a monumental task that it doesn't seem realistic, they're too far down the path of generalized autonomy
Agreed. But imagine how many fewer tens of thousands of people who paid $8k and $10k for "FSD" would have paid, if the narrative was:
"one day, not sure when, and not sure if this hardware will even allow it, but its possible that we could live to see a robotaxi service. Its going to be many years out, but its something we can aim for".

So instead, even though you absolutely know its not true, you make up a date thats 12-18 months away and say that the Robotaxi service WILL be a thing. BUt you add in a few "most likely" or "probably" in your tweet to cover yourself legally.

and voila..hundreds of millions of dollars paid to your company via FSD purchases.
 
1642803144308.png
 
It is possible that FSD Beta is the testing of the autonomous stack but that the public release of "autosteer on city streets" will be a stripped down version of FSD Beta. So, Tesla could release the parts of FSD Beta, as "autosteer on city streets", that they feel are reliable enough for full public release. It would allow them to say that they fulfilled the promise of releasing "autosteer on city streets" while they are still working to make FSD Beta reliable enough.

I can kind of see why Tesla calls it "autosteer on city streets". I think the idea is that "autosteer" does more than just steer to stay in the lane, "autosteer" can now turn the wheel all the way to 90 degrees to make turns at intersections. Additionally, "autosteer" is capable of steering to stay in the lane on residential streets with no lane lines or steer around parked cars.

Personally, I think of FSD Beta as "City NOA" as the main feature seems to be that it tries to navigate a route to get you to a destination that you put in. It does auto lane changes to stay on route, auto lane changes to pass slow traffic and makes turns at intersections to follow a route which is similar to what NOA does, except it does it for city streets instead of highways.

Interestingly, "Autosteer on city streets" used to be called "City NOA" back in 2019 when Tesla added a list of features to the FSD page. Personally, I feel like "autosteer on city streets" sounds less autonomous than "City NOA". It seems that Tesla might have changed the name to downplay the autonomy of the feature, perhaps to avoid criticism when the feature is released and it falls short of L4/L5 autonomy.

I used to be really convinced that this was exactly how it would go down. But, the development work they've been doing extends well past even the City NoA.

Things like the Single Stack which will impact both city NoA, and highwat.

Heck the first thing that happened when I got FSD Beta was I got stuck with Tesla Vision.

In a way I prefer what they're doing because I honestly was never all that happy with TACC/AP/NoA. So I don't mind that they're taking a bulldozer to it for those of us with HW3 vehicles.

With that being said I do expect Tesla to put restrictions on FSD Beta when it goes to a general release. But, it likely won't have the degree of limitations I was concerned with.

Like I expect it to fully handle 4 way stops where it makes a decision itself when to go. My original theory was they would always force the driver to confirm. Like it would beep and you'd press to go.

The other thing could impact Tesla's decision is Ultracruise. I think Tesla will try to take a position where it will be more flexible than a competitors.
 
  • Like
Reactions: pilotSteve
Agreed. But imagine how many fewer tens of thousands of people who paid $8k and $10k for "FSD" would have paid, if the narrative was:
"one day, not sure when, and not sure if this hardware will even allow it, but its possible that we could live to see a robotaxi service. Its going to be many years out, but its something we can aim for".

So instead, even though you absolutely know its not true, you make up a date thats 12-18 months away and say that the Robotaxi service WILL be a thing. BUt you add in a few "most likely" or "probably" in your tweet to cover yourself legally.

and voila..hundreds of millions of dollars paid to your company via FSD purchases.
Tesla isn’t making up arbitrary dates. Elon truly believes the dates they announce are achievable. Being late or wrong doesn’t make them charlatans.
 
Tesla isn’t making up arbitrary dates. Elon truly believes the dates they announce are achievable. Being late or wrong doesn’t make them charlatans.
It makes them something. Elon has been consistently wrong on FSD delivery dates for 5+ years now. He’s either lying, or simply so completely out of touch that he has no idea what’s going on in his own software development group. There are really no other options.
 
It makes them something. Elon has been consistently wrong on FSD delivery dates for 5+ years now. He’s either lying, or simply so completely out of touch that he has no idea what’s going on in his own software development group. There are really no other options.
Have you considered it’s a ginuinely hard problem that’s easy to underestimate? Look at the evidence. No company in the world has achieved general FSD so maybe.. just maybe… it’s VERY hard and the estimates were simply too optimistic.
 
  • Like
Reactions: SucreTease
Have you considered it’s a ginuinely hard problem that’s easy to underestimate? Look at the evidence. No company in the world has achieved general FSD so maybe.. just maybe… it’s VERY hard and the estimates were simply too optimistic.
No company in the world is claiming L5 next year for 6 straight years.
How is it that you can't understand what @dramsey is saying?