Coast to coast drive happening this year for all FSD Teslas!

[diplomat33]

Yeah, strange how they wouldn't let anyone record their demo rides though. All we got was that one Telsa produced video which for all we know could have been run 97 where it finally managed to go the whole loop without crapping out.

No, there were plenty of testimonials from the test drives, on youtube, twitter and on various websites.

Here's one I posted awhile a back in the original autonomy investor day thread from Morgan Stanley but there are many more.

 

[juliusa]

That idea has been put on hold indefinitely.

How is it going to drive coast to coast without charging? I thought the car was going to drive itself cross-country, and the person was there just for compliance with laws.

 

[banned-66611]

No, there were plenty of testimonials from the test drives, on youtube, twitter and on various websites.

Here's one I posted awhile a back in the original autonomy investor day thread from Morgan Stanley but there are many more.

Right, so we have a failure rate of 1 per 20 minutes, on a known and well tested route, and the failure was it couldn't follow the planned route.

That sounds like they are a hell of a long way from having something that approaches "full self driving", and light years away from what Musk just promised we would see THIS YEAR.

 

[diplomat33]

Right, so we have a failure rate of 1 per 20 minutes, on a known and well tested route, and the failure was it couldn't follow the planned route.

That sounds like they are a hell of a long way from having something that approaches "full self driving", and light years away from what Musk just promised we would see THIS YEAR.

That is a flawed argument. You are taking a small sample and making an over generalized conclusion. It was one trip that happened to have 1 fail in it. There were other test drives with no fails at all. I could show you a video of a Waymo car failing on just a 5 mile trip but that doesn't mean that Waymo cars fail every 5 miles.

Look, if you want to continue to doubt and be skeptical, that's your prerogative of course. Tesla has given us some concrete evidence that they are making good progress on FSD. We will probably get "feature complete" software by end of this year, if Tesla stays on track with their timeline.

 

[CK_Stuggi]

With all due respect. Thank you for the Morgan Stanley report, I missed that.
Not having been there first hand though, I remain not regarding this as concrete evidence.
Either way; the year has 32½ weeks left, since AID 3½ weeks have passed without a major new feature rolled out. Time will tell.
Certainly would love to hear the upgrade schedule to HW3 and my "upgrade" to "official Beta tester". Hmm - no news there either...

 

[conman]

How is it going to drive coast to coast without charging? I thought the car was going to drive itself cross-country, and the person was there just for compliance with laws.

The answer is... it's not. That sort of coast to coast drive was given up on and they changed the bar.

 

[MikeS1]

The ignorance and narrow-minded people on here is astonishing. The system can get exponentially better with NN training.

How are so many people clueless if this fact? The FSD we see in the fall with be waaaay better than what was shown last month.

 

[diplomat33]

How is it going to drive coast to coast without charging? I thought the car was going to drive itself cross-country, and the person was there just for compliance with laws.

The answer is... it's not. That sort of coast to coast drive was given up on and they changed the bar.

It certainly appears that Tesla has put the automated snake charger on the back burner or possibly even nixed the whole thing completely. We got that one prototype video like 3 years and then not a peep about it since. If it were a high priority, I am sure Tesla or Elon would be yapping about it constantly.

Having said that, you can still do a legit coast to coast FSD drive with no human intervention, even without the snake charger. It just means having a human passenger who gets out of the car to plug in the charger or have a human present at the charger to plug in when the car arrives. The car could still do all the driving, including pulling up to the charger with no human intervention.

The real issue is how do you robotaxis and the Tesla Network without automated chargers. Obviously, if you want your robotaxi to be able to recharge when there is no human in the car, it would need some type of automated charging system. And charging would need to happen outside of operating hours since you don't want passengers to have to wait while the car is charging. So having passengers get out to charge the robotaxi is out of the question. The only way it could work without automated charging is if you restricted the operating hours to only when the car still has enough charge. So the robotaxi might leave it's home base with a full battery, run around town picking up passengers, and then when the robotaxi is almost at an empty state of charge, it would stop accepting more passengers and return to home base where a human would plug it in. So the Tesla Network could work without automated chargers at least in the beginning.

My guess is that automated chargers will still happen but are not a top priority. Obviously, FSD is the first priority. Automated chargers can wait until Tesla is closer to achieving robotaxis.

The ignorance and narrow-minded people on here is astonishing. The system can get exponentially better with NN training.

How are so many people clueless if this fact? The FSD we see in the fall with be waaaay better than what was shown last month.

Exactly! Some folks are guilty of fixed thinking. They are only seeing FSD by what Autopilot can do now. But that is a mistake because it ignores the fact that FSD is not fixed; it's in active development now and Tesla is making great strides every day.

 

[iTech]

The ignorance and narrow-minded people on here is astonishing. The system can get exponentially better with NN training.

How are so many people clueless if this fact? The FSD we see in the fall with be waaaay better than what was shown last month.

That's all well and true...if you don't add a deadline to your statement. Tesla did, that's the problem.

 

[Wooloomooloo]

The ignorance and narrow-minded people on here is astonishing.

Skepticism isn't narrow minded and ignorant. Cynicism is, and so is the willingness to believe anything (on opposite ends of an extreme). I'm living and breathing ML in the NLP world every day, and I know one thing for sure. It ain't easy.

The system can get exponentially better with NN training.

Can it? Actually mathematically, NN's may require exponentially greater data to become marginally better, depending on where you are on the curve.

How are so many people clueless if this fact? The FSD we see in the fall with be waaaay better than what was shown last month.

Maybe, and we all hope so. But is "waaaaaay" better in line with what has been promised, or something more tangible? When Tesla went from 200 Model 3's per week to 1,000 Model 3's per week, that was "waaaaaay" better, but it wasn't 6,000 per week, which was promised.

There are two conversations going on in this thread. One is "come on, this is progressing and is really cool", which I can buy. The other is "hang on, Tesla are not going to have full autonomy in 2019", which I can also buy. Are you saying that those of us taking that latter position are "narrow minded and ignorant" or "clueless".

So back to my job. It sure ain't easy, but it can be easier with great expectation management. Fact, fiction and fantasy is a wonderful way of framing it.

 

[Wooloomooloo]

That's all well and true...if you don't add a deadline to your statement. Tesla did, that's the problem.

Oh you beat me to it!

 

[Eno Deb]

That's all well and true...if you don't add a deadline to your statement. Tesla did, that's the problem.

No problem. Elon has proven many times that he can bend time.

 

[Electroman]

I am bullish on Tesla autonomy. I believe their approach will get them to FSD ahead of others.

I also firmly believe they will not have FSD by end of 2020. I am fine with that.

 

[willow_hiller]

I am bullish on Tesla autonomy. I believe their approach will get them to FSD ahead of others.

I also firmly believe they will not have FSD by end of 2020. I am fine with that.

A very reasonable position. Do you have any particular reason why you think FSD won't come until the end of 2020? Regulations? More difficult than Elon thinks?

 

[Daniel in SD]

A very reasonable position. Do you have any particular reason why you think FSD won't come until the end of 2020? Regulations? More difficult than Elon thinks?

No one has done it despite many other companies working on it for about a decade with $80 billion of investment. It's a really difficult problem.

 

[RDoc]

The ignorance and narrow-minded people on here is astonishing. The system can get exponentially better with NN training.

How are so many people clueless if this fact? The FSD we see in the fall with be waaaay better than what was shown last month.

KoolAid is soooo delicious, especially with that little touch of bitter almond flavor. Yumm.

 

[banned-66611]

That is a flawed argument. You are taking a small sample and making an over generalized conclusion.

That was my point, actually. This is all Tesla has given us. The very best they could do, in a carefully managed press event.

In other words that's the best case, the real general performance is going to be worse.

 

[diplomat33]

That was my point, actually. This is all Tesla has given us. The very best they could do, in a carefully managed press event.

In other words that's the best case, the real general performance is going to be worse.

Nope. You are just picking the 1 test drive with a fail. You are picking the worst case, not the best case. Tesla did lots of test drives that day with 0 fails. So that test drive with the fail was the worst example, not the best example. And FSD will improve. The real FSD we get will be much much better than the test drives.

 

[banned-66611]

Nope. You are just picking the 1 test drive with a fail.

No, Tesla are picking it. If they had multiple successful test drives they would show evidence of that, and have allowed journalists to film their rides.

The onus is on Tesla to prove they can do more, otherwise we must assume that what they did allow to be published was the absolute best case for them.

 

[willow_hiller]

No, Tesla are picking it. If they had multiple successful test drives they would show evidence of that, and have allowed journalists to film their rides.

The onus is on Tesla to prove they can do more, otherwise we must assume that what they did allow to be published was the absolute best case for them.

Dozens of investors published testimonials about the test drives. You think if there was an NDA, news of it would have leaked by now.