Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

FSD Beta Videos (and questions for FSD Beta drivers)

This site may earn commission on affiliate links.
With all due respect to your opinion I'd like to confirm that you're aware that it's a parody of the CNN hit piece, which I should have mentioned. Here are three videos
1) The original CNN video
2) The video shot by the vehicle owner from the back seat during the same (CNN) drive
3) The parody of the CNN video.

I would argue that the original CNN video was so distorted it was almost a parody on it's own.

Count me as having a childish sense of humor, too, then. I thought it was hilarious.

SNL level material.
 
With all due respect to your opinion I'd like to confirm that you're aware that it's a parody of the CNN hit piece, which I should have mentioned. Here are three videos
1) The original CNN video
2) The video shot by the vehicle owner from the back seat during the same (CNN) drive
3) The parody of the CNN video.

I would argue that the original CNN video was so distorted it was almost a parody on it's own.

Thanks, I know it was a parody of the CNN report. I don't think it was something to make a parody of. CNN highlighted the worst of FSD, and that is status quo sadly.
 
CNN highlighted the worst of FSD, and that is status quo sadly.
There's certainly nothing wrong with showing an honest evaluation of what FSD can and can not do, but IMO, that wasn't it.

It was an over dramatized piece with an overwhelmingly bearish tone.

I honestly wouldn't mind if they did a piece that cherry picked out the bad stuff, as long as they kept the drama toned down.

But that piece was an embarrassment.

I can just see their next Tesla article...

A Tesla has a small dent on the front fender:

"This is CNN coming to you live at the scene of a horrific crash involving this Tesla. As you can see, the car is a total loss, and at this time we have no word on if there are any survivors."
 
BTW the car owner that made the "real truth" is in my view a joke. The car did try to crash 4 times, that is the fact. Why would he try to hide the facts? What is his hidden interests?
Why are Tesla fans annoyed by CNN? Why are they not annoyed by the FSD system trying to crash every other mile?
 
  • Disagree
Reactions: Terminator857
BTW the car owner that made the "real truth" is in my view a joke. The car did try to crash 4 times, that is the fact. Why would he try to hide the facts? What is his hidden interests?
Why are Tesla fans annoyed by CNN? Why are they not annoyed by the FSD system trying to crash every other mile?

I don't really care what CNN does, or CNBC, or anyone else. So many negative pieces have come before this. Did any of them make a difference? No.
 
Last edited:
Checkout the AI day. You will get a better understanding of what parts are NN and how the cost-optimization path planner works.
Definitely worth a rewatch, now that I've had firsthand experience with the beta! Is it your understanding that the NN is completely in charge of the red light (mis)behavior seen in the video? Also wondering if perhaps the lane guides within the intersection confused the network and made it think it was stopping behind the line? Still a mistake that no human driver would ever make.
Currently FSD Beta is Vision + Map based. But map is needed to figure out turn lanes, turn restrictions etc. FSD doesn't "read" those - just as it doesn't read road names. If there are map issues or FSD can't map what it says correctly with the 2-D map information, we get weird issues.
Sure, and this explains most of the failures in the video. But FSD will absolutely need to be able to dynamically read road signs and markings in order to reach any sort of city-street L4, or even halfway useful L2. I wonder where Tesla is at internally with this.

A totally different approach they could take would be to make a monolithic GAN-style network that uses unsupervised unlabeled learning to simply predict what a human would do in any given situation, with an adversarial network able to take FSD-predicted paths and actual human-driven paths and determine which is which, then train the FSD network to generate paths that fool the discriminator. With this approach it could learn all the road signs implicitly, with no human labeling or annotation, and it might even be more robust than feature-engineered supervised learning.
 
Why was the CNN guy sitting so close ?

Personally I'd not let any "journalist" touch my 3 to put out hit pieces to help the oil barons.
To use more stock-market terminology here, I think any company coming up with good autonomous driving or driver assist systems would be extremely bullish for all energy demand.

Imagine how many more miles would be driven annually if cars could just drive themselves flawlessly. Road trips and long hauls would skyrocket, people who aren't able to drive for whatever reason would now be on the road, etc etc. Even if limited to EVs, which would surely only apply at the front end before trickling into all types of vehicles (especially if the goal is reduce accidents now), it would all feed into the overall demand for energy and would end up benefitting oil & gas both directly in ICE cars and indirectly through increased demand for power through the grid (still heavily reliant on fossil fuels), more wear and tear on components made out of oil & gas, and so on.
 
Yesterday wasn't the best day to go for a ride, so I waited until Thanksgiving morning to give FSD beta 10.5 a roll. FSD beta is timid (FSD profile is "Average"), but overall the car did better than I expected. Towards the end you can see it really doesn't like roads with no painted lines. Overall, it still did ok on the cobblestones in Manhattan (although steering was very lively), with all the randomness of NYC moving around it.

 
... Why are they not annoyed by the FSD system trying to crash every other mile?
Interesting we have such a wide difference in opinions. On one side we have whole mars catalog and powertood saying it the best thing since sliced bread and hours of no disengagements drives. Then we have others saying it tries to kill you every other mile. Is FSD that much better in the bay area?
 
Interesting we have such a wide difference in opinions. On one side we have whole mars catalog and powertood saying it the best thing since sliced bread and hours of no disengagements drives. Then we have others saying it tries to kill you every other mile. Is FSD that much better in the bay area?
Daktari is in Norway.

Of course, people who say no disengagements push FSD farthest and have learned to live with its quirks. They also know when to use the go pedal to help FSD. I think I can also do a lot of no disengagement drives if I figure out the right go pedal intervention strategy for hesitant round abouts and turns.

It’s possible Bay Area has fewer map issues and that helps greatly.
 
Interesting we have such a wide difference in opinions. On one side we have whole mars catalog and powertood saying it the best thing since sliced bread and hours of no disengagements drives. Then we have others saying it tries to kill you every other mile. Is FSD that much better in the bay area?
To CA's credit stuff generally seems to be well marked. And of course Tesla engineers have driven those roads far more as well, so they may as well be full high res mapped.

FSD cannot handle my NYC street whatsoever, but I get it - it's a wide one-way street that people treat as 1, 2 or 3 lanes sometimes. People double park. There are no lane lines. FSD causes the car to swing wildly left and right as it can't decide whether the street is actually 2 lanes or 1 lane. And to be completely honest, nobody seems to know whether it's a 2 lane street or not, myself included.
 
  • Like
Reactions: DanCar
To CA's credit stuff generally seems to be well marked. And of course Tesla engineers have driven those roads far more as well, so they may as well be full high res mapped.
Nothing like "full high res mapped". Probably there are fewer errors - and whatever errors Tesla internal testers found would have been corrected.

FSD cannot handle my NYC street whatsoever, but I get it - it's a wide one-way street that people treat as 1, 2 or 3 lanes sometimes. People double park. There are no lane lines. FSD causes the car to swing wildly left and right as it can't decide whether the street is actually 2 lanes or 1 lane. And to be completely honest, nobody seems to know whether it's a 2 lane street or not, myself included.
FSD handles double parking quite well. It doesn't really care whether there is double or octo parking. Just goes around the obstacle. If multiple drivers are driving at the same time next to each other in a single one-way lane ... that could be an issue.
 
Ofcourse they did. The SP was suppressed for so long and shorts made billions.

Until the truth could not be suppressed and they lost billions, and lots more people became teslanairres.

This is a good summary. I actually didn't know that so many of the prominent short sellers had such a significant decrease of assets under management. They've been destroyed.

 
FSD handles double parking quite well. It doesn't really care whether there is double or octo parking. Just goes around the obstacle. If multiple drivers are driving at the same time next to each other in a single one-way lane ... that could be an issue.
In my case, it just sat behind this line of cars on the right. The car actually started on the left, ping ponged left and right, but then saw cars lined up and moved over to the right to get in line.

This street is definitely an edge case - you can see how wide it is. FSD is clearly looking for anything that might make sense to be a lane. It may do better when there's more traffic to give the car some context clues. When it's empty like this, it has no idea.

Screen Shot 2021-11-25 at 8.32.17 PM.png
 
  • Informative
Reactions: Phlier
In my case, it just sat behind this line of cars on the right. The car actually started on the left, ping ponged left and right, but then saw cars lined up and moved over to the right to get in line.

This street is definitely an edge case - you can see how wide it is. FSD is clearly looking for anything that might make sense to be a lane. It may do better when there's more traffic to give the car some context clues. When it's empty like this, it has no idea.

View attachment 737446
Not familiar with NYC convention, but is this road officially a 2 lane road? We do have certain roads here in SF that have super wide lanes, but the only time when there may be two cars next to each other in the same lane is if one is making a right turn or a left turn.

If this is a single lane road officially, it may be assuming those cars are going forward and not turning right or stopped (waiting for passengers etc), so it would "make sense" to get in line.
 
Not familiar with NYC convention, but is this road officially a 2 lane road? We do have certain roads here in SF that have super wide lanes, but the only time when there may be two cars next to each other in the same lane is if one is making a right turn or a left turn.

If this is a single lane road officially, it may be assuming those cars are going forward and not turning right or stopped (waiting for passengers etc), so it would "make sense" to get in line.
lol, this is the question I've even asked NYC officials many times. I've lived here for 20 yrs, and every few years, I try and find out because someone will do something crazy like 4 across and it'll bother me. Usually I see 2, and I assume one is for turning as you suggest. However, nobody will officially claim it is two or one lane, probably because it would mean they would own the problem and be forced to put lines or something.

I don't care much either way - it's been like that for longer than I've lived here. Just amuses me. I'll keep reporting it in FSD beta and see if maybe the car figures it out eventually.