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

Poll: How will youtubers rate FSD Beta 9?

How will youtubers rate FSD Beta 9?

  • Best thing since sliced bread

    Votes: 8 18.2%
  • Second coming of Lord Elon

    Votes: 7 15.9%
  • You haven't lived until you've experience FSD Beta 9

    Votes: 12 27.3%
  • Meh

    Votes: 3 6.8%
  • Needs work

    Votes: 7 15.9%
  • Much better than FSD Beta 8

    Votes: 18 40.9%
  • Almost there

    Votes: 11 25.0%
  • Ready to deploy: launch! launch!! launch!!!

    Votes: 3 6.8%
  • Good in most cases, dangerous in some cases

    Votes: 7 15.9%
  • Do we care?

    Votes: 4 9.1%
  • Other, please comment

    Votes: 1 2.3%

  • Total voters
    44
This site may earn commission on affiliate links.
How will youtubers rate FSD Beta 9?

Probably depends on the youtuber. Some will say "Mind blown! Elon has solved L5!" Others will say it needs some work. It will probably depend too on where and when they use it. If they use on easy routes, it will probably perform better, leading them to say it is amazing. If they use in situations that require disengagements, they will probably say it needs work.
 
  • Like
Reactions: DanCar
The inputs to the system are so random and the performance is so inconsistent that youtubers will be able to conclude whatever they want.
Keep in mind that Elon said V9 is a "foundational" change so there may be no initial improvement in performance.
Does their agreement with Tesla even allow them to objectively measure performance? It would be interesting to measure disengagements every day on a fixed commute.
Tesla Owners Silicon Valley and Whole Mars Catalog will definitely say it blows their minds (even while running over a curb :p ).
 
The inputs to the system are so random and the performance is so inconsistent...
I would say that: It will be highly consistent, highly reproducible in certain locations such as some sharp turns even for V9

The China video below is with traditional firmware (non-FSD beta) and both its map on the right and its instrument clusters clearly show that the road is curvy but the car still drove in a straight line and crashed recently.


1625866285030.png


I suspect if you drive the new FSD beta 9 in this location, you'll get the same result!

Ok. China is too far, how about drive your FSD beta 9 in the Oakland, CA area with the same curvy road scenario? It was done in FSD Beta 8.2 and the driver had to take over to prevent from crashing into the iron fence. Notice the system clearly draws the red sharp 90-degre turn road shoulders.


1621724144938-png.665110
 
Last edited:
Teslatino:
  1. Awesome view
  2. Getting into lane much earlier, much better.
  3. Path prediction looks really really good
  4. A little bit slow, that is o.k.
  5. Got into left lane to late for left turn. In congested traffic situation that would not have been good.
  6. Didn't handle a dual left turn lane well. Crossed lane. 5:00 into the video
  7. Feels more assertive. At least now it got into the lane on time. 6:00
  8. 7:00 Didn't handle a police block. Manually took over.
  9. 9:15 Mind blowing that it is only using cameras.
  10. 11 minutes: looks really nice, I like it.
  11. Looking like a very good ride.
  12. 11:50 Why did it move into the right lane. No traffic. Right turn 3 miles ahead.
  13. 12:40 Still feels like beta . Recieved a legend what every color means.
  14. 13 minutes, display looks awesome.
  15. 13:15 decided to change lanes again. Doesn't know why.
  16. 14:45 No phantom braking
  17. 15:00 Changed again to middle lane near intersection.
  18. 16:30 failed to get next to charging stations, just came close.
  19. 18:25 A little bit slow
  20. 19:30 Feels more smooth than previous version
  21. 21 minute So far hasn't screamed at me for not paying attention
  22. 22:15 I'm very happy with this
  23. 23:00 It is doing a very very good job with no traffic
  24. 23:35 A little bit of phantom breaking with yellow lights
  25. 25:00 More phantom breaking with yellow caution lights.
  26. 27 minutes Consistently moves into the middle lane when approaching intersection
 
  • Informative
Reactions: diplomat33
To be honest, after 3 hours of FSD 9 videos, only a cynic would not be impressed.


There are a lot of cynics though!


You have cars on the road driving themselves around with the primary input - 5 year old webcams - as some "experts" on here have put it.
I view it as doing a useless impressive thing much more cheaply. :p I just don't see much value in this sort of partial autonomy. It seems unlikely that they'll be able to get to real autonomy with the current hardware when so many other companies have invested even more money and engineering resources and achieved only extremely limited FSD.
Autonomy day was April 2019. Can we quantify the progress since then? How much further do they have to go?
 
To be honest, after 3 hours of FSD 9 videos, only a cynic would not be impressed.


There are a lot of cynics though!


You have cars on the road driving themselves around with the primary input - 5 year old webcams - as some "experts" on here have put it.
Yes it is very impressive from a certain point of view. It is not impressive from the expectations that Elon has set up such as cars driving themselves last year. These cars are no where near driving themselves without supervision.
 
I view it as doing a useless impressive thing much more cheaply. :p
Is TACC useless? Why Level 2 FSD better or worse than TACC?

I just don't see much value in this sort of partial autonomy. It seems unlikely that they'll be able to get to real autonomy with the current hardware when so many other companies have invested even more money and engineering resources and achieved only extremely limited FSD.
Autonomy day was April 2019. Can we quantify the progress since then?
I don't think it is mostly a hardware issue. Once general A.I. is developed it will be able to drive fine with current sensors. Of course it will be able to drive better under harsh conditions with more sensors. Keeping it simple until the software catches up is a good strategy in my opinion.
How much further do they have to go?
Minimum five years is my guestimate. Could be 10+ years. Although I will truly enjoy level 2 FSD long before that.
 
Is TACC useless? Why Level 2 FSD better or worse than TACC?
Because all I have to do to monitor TACC is look straight ahead.
How do you monitor a right turn on red? Do you look left or right?
I don't think it is mostly a hardware issue. Once general A.I. is developed it will be able to drive fine with current sensors. Of course it will be able to drive better under harsh conditions with more sensors. Keeping it simple until the software catches up is a good strategy in my opinion.
I agree that the A.I. issue is way more challenging than perception. Other companies are trying to get around that by using remote assistance. I think selling autonomy for current hardware was a huge mistake. Using the fleet for data collection is good but I think they'll regret being locked in to a hardware platform for development.
Minimum five years is my guestimate. Could be 10+ years. Although I will truly enjoy level 2 FSD long before that.
I'm on record saying they will never release L2 FSD :p
Though I see that they've added some confirmation to the beta so I'm probably going to be proven wrong. Exciting times!
 
  • Like
Reactions: Terminator857