Discuss: New 'Wile E. Coyote' Wall Video Puts Tesla's FSD V13 to the Test

[Not a Tesla App]

Tesla Owner challenges Mark Ruber's Wile-E-Coyote Video, demonstrates FSD V13 and AI4 is capable of stopping for a fake wall - HW3 needs improvements.… Read More

New 'Wile E. Coyote' Wall Video Puts Tesla's FSD V13 to the Test

Tesla Owner challenges Mark Ruber's Wile-E-Coyote Video, demonstrates FSD V13 and AI4 is capable of stopping for a fake wall - HW3 needs improvements.

www.notateslaapp.com

 

[PrescottAZRichard]

I'm impressed the CT saw the way TBH. You can see there's a line across the road where the wall is on the visualizations. Still, it's best to stay on the good side of your neighborhood coyotes and roadrunners.

 

[K.I.T.T.]

So, the facts, unbiased.

Mark Rober comes up with meticulously engineered tests and walls and fools autopilot. But does not try FSD. Damn.

Other guy puts together something way less meticulous - strangely when CyberTruck was engaged, wrinkles in the wall wouldn't have fooled the Coyote, not to mention Roadrunner. Not to mention that he barely exceeds 30 MPH and doesn't test mannequins nor water walls.
So we don't even know if that's due to HW4. Even if it was, you all having cars with lesser HW versions should be rightfully concerned, as I should be with my autopilot HW4.

Or perhaps not - HW4 with autopilot wasn't tested, so we don't know if it's the hardware's or the software's fault.

Now I expect Marc Rober to come up with a comprehensive follow-up to respond to all the attacks, compare a matrix of HW3/4 and autopilot/FSD combinations.

Unbiased analysis: it's just logic. The choice of getting rid of other sensors than vision is only coupled with the intentions of using a single "FSD" stack for Optimus and self driving cars on Mars, as I was anticipating a year ago. Now that the intentions to launch a Starship with a team of Optimus has confirmed officially, who can doubt that we are just hamsters in Musk's FSD wheel for the Mars master plan?

 

[PrescottAZRichard]

So, the facts, unbiased.

Mark Rober comes up with meticulously engineered tests and walls and fools autopilot. But does not try FSD. Damn.

Other guy puts together something way less meticulous - strangely when CyberTruck was engaged, wrinkles in the wall wouldn't have fooled the Coyote, not to mention Roadrunner. Not to mention that he barely exceeds 30 MPH and doesn't test mannequins nor water walls.
So we don't even know if that's due to HW4. Even if it was, you all having cars with lesser HW versions should be rightfully concerned, as I should be with my autopilot HW4.

Or perhaps not - HW4 with autopilot wasn't tested, so we don't know if it's the hardware's or the software's fault.

Now I expect Marc Rober to come up with a comprehensive follow-up to respond to all the attacks, compare a matrix of HW3/4 and autopilot/FSD combinations.

Unbiased analysis: it's just logic. The choice of getting rid of other sensors than vision is only coupled with the intentions of using a single "FSD" stack for Optimus and self driving cars on Mars, as I was anticipating a year ago. Now that the intentions to launch a Starship with a team of Optimus has confirmed officially, who can doubt that we are just hamsters in Musk's FSD wheel for the Mars master plan?

So you were right. You are right. You continue to be right.