Caveat: a generous understanding of “Elon time” aside, anyone else see this on Twitter today? Elon asked about the autonomous drive from LA to NYC, and he not only says this year but adds this
“Are their still plans to drive from NYC to LA on full autopilot?”
“We could have gamed an LA/NY Autopilot journey last year, but when we do it this year, everyone with Tesla Full Self-Driving will be able to do it too”
Again, “Elon time” aside, this is exciting!
Of course, watching the latest FSD video by Tesla, it doesn’t seem that much of a stretch really.
So, I need to drive from RI to Sacramento. I drive from RI to NYC. Then FSD can take me to LA. Then I drive to Sacramento.
Okay, guilty of being sarcastic. But my point is whether NYC-to-LA is a carefully vetted route, or whether FSD will be ready in a more general sense.
lol He's BSing. No way this will happen even in the next 5 years.
As long as every road you use has painted lines on both sides. You hit a small section where a line is not there on both sides AP will switch off. Amateurs.
We missed you, Hal. So great to hear from you again.
If that’s the goal it means a lot of shorter more manageable segments will become a reality. I’m all for that. Exciting times ahead.
Hal first of all you know damn well there is a huge difference between fsd and autopilot.
But to further debunk your shit, I have driven on autopilot on a country road that discontinued its paint. No paint on either side, and no center lane markings. Of course I was very careful and highly alert. Autopilot does not disengage, and it stayed on the right side of the road for a half mile. Once I manually terminated autopilot, I could not re-engage it. But you’re just dead wrong about AP switching off.
I second the no lines rebuttal, my autopilot doesn't disengage. Sometimes I will get a warning to take control of wheel but it usually remains in control in these situations.
It still lines my car up with the center of the town's double rotary. Impress me, navigate that first...
None of us has any idea on what timeline this will be possible. We cannot use our experience with the current version of autopilot running on our current hardware.
I'm not saying I believe or disbelieve Elon's timeline. I'm just saying that everyone who says with such certainty that we are years away has no basis for the statement other than skepticism.
We'll have a better idea when they start sending beta versions of the new neural net to cars with the new hardware. Until then, any current experience we have is useless. It's like saying, BEV's will never work because my ICE vehicle can't drive on its 12 volt battery.
@ spuzzz123 & @kevin_rf
Uhm.. you both know that Hal has been and will always be full of shit and nonsense. Others are seeing it too.
Furthering what spuzzz123 said...you need to carefully parse what Elon said... He said it would be an "Autopilot" journey. Now you can mostly do that now if by Autopilot you mean TACC and Autosteer. I would venture that he means NoAP but I really don't think you can go so far as to say that he meant Level 3/4 autonomous driving. Him saying everyone with FSD would be able to do it doesn't really mean much because since EAP moved into the FSD set you can't say that FSD explicitly means Level 3/4/5 autonomy.
I assume it means something akin to what we saw the model 3 do in their latest FSD video.
@CharleyBC - "...my point is whether NYC-to-LA is a carefully vetted route, or whether FSD will be ready in a more general sense."
I think what Elon was saying is that up until now they could have done it on a vetted route but that now he's saying any route will be possible when they do it later this year.
@Hal Fisher - "As long as every road you use has painted lines on both sides. You hit a small section where a line is not there on both sides AP will switch off."
Except that that isn't how AP usually works in real life and were talking about FSD not AP, and we're talking about features that haven't been released yet. But hey, you can make up some more stupid stuff and post that and see if anyone believes you then.
Apparently not everyone watched the autonomy day presentation.
@Bighorn, right? That was a fascinating presentation, the second guy was my fave. That presentation revealed just how far far along they really are.
The future is soon. I can hardly contain my excitement.
The jury is still out as to which system will produce the most reliable and therefore first to the mass market Level 5 autonomous driving at an affordable price. Tesla has some encouraging statements recently at their event, but remember who the audience was... people who Tesla desperately needs money from.
has = had
Res ipsa, do all attys make reference to the jury being out?
Only litigators I think...
I regularly drive from downtown St. Petersburg on US 275 through an automated toll lane, over the Skyway bridge, and down US 75 to Sarasota without any input. Why couldn't you drive most of the way from NY to LA right now?
People need to stop using their understanding of AP2.5 to judge what Elon might be able to demo on HW3 + APx.x
I don't know whether he'll be able to pull something of by year end, but speculating is pointless
I want HW3 upgrade please!
@Hal Fisher, you are definitely wrong. I live in area where there are plenty of areas where painting stops, yet if I am using AP it continues to work until I get to point where I have to stop car for Stop Sign or Red light. I have only started this testing since Tesla added Red Light detection to car when running on AP. Now I will agree even though the car has marker for where lines should be, I agree if AP is not enabled prior on a road without actual painted lines it will not let you engage AP.
@efuseakay, in 5 Years multiple manufactures including Tesla will have release fully autonomous cars. Matter of fact most Manufactures working on Full Self Drive fully expect it to be Level 4 possibly Level 5 between 2020 and 2023. Just so you know the difference in Level 4 and Level 5 is wording goes from Most Road and Road Conditions to All Roads and Road Conditions. My prediction within 7 years the Insurance industry will make it too expensive for the average driver to actually be able to afford insurance on a normal non self driving car.
Coast to Coast FSD could definitely happen this year! Autonomy Investor Day, FSD test drives and FSD video show that Tesla has what it takes now to do it.
The FSD deniers remind me of flat earthers.
Snakey plug for charging too?
Don't even need to leave the vehicle except for bio breaks. You could always hang it out the window I guess.
As long as it is on a straight road that is well marked it might be possible. However I can't even drive 20 miles on the highway without something happening and it switching off, let alone driving on side roads where traffic lines go perpendicular for cars in the opposite lane making a left hand turn. Also I can't imagine coming to a four-way stop sign or flashing red light and having the car know which other car is going to move first or has the right of way.
Today I was on I-95 and put on the turn signal to go into the HOV lane. Since it was a double dotted line the car freaked out since it didn't know what to do after it crossed the first dotted line even though they are only about a foot from one another . When going out of the HOV lane into the next Lane it has the same problem recognizing what to do with two dotted lines being right next to each other.
Some states have toll booths where the lines disappear before you get to them and you have to figure out what Lane to get into to pay the toll. Also some lanes are exact change and some are only work if you have a transponder which lets you drive through. How is it going to know which one to go into since they are not standardized and sometimes lanes are closed and if there is heavy traffic is very difficult to get into the lane that you need to be in. I would love to see them try and get a car through the Lincoln Tunnel or Holland Tunnel during rush hour when everybody is riding each other's bumper and going from about eight Lanes into only a few and you are only inches away from the cars along side of you that have three or four cars taking up the space of only two lanes. or four cars in one or two lanes
I’m not wrong at all and the story actually gets worse. I drive the commute lane on i5 north between san diego and oceanside workdays. Coming home one night on i5 north right after the commute lane ends 3 cars had hit a huge pot hole in the commute lane, two of them were teslas with flat tires. So they can’t detect pot holes. So The next day they had the holes (I don’t know how two happened at once, maybe a rim cause the second hole?) filled with about 8 ft of asphalt but they didnt paint the line over the patch which is on the left side on the lane. Well if im driving 10 mph or so I can pass over the area fast enough that FSD doesn’t switch off, but if I’m going much slower fsd will turn off. Makes no sense why it would shut for any other reason. I think they should stay next to which ever line that’s the straightest, that way the don’t try to straddle a merge lane so much. Odd is they seem to get it right when it comes to exits, it stays on the left lane. Maybe it just freaks if a left lane marker disappears?
Meant to say it stays on the left ‘line’.
Just getting all of the cars upgraded from HW2/2.5 to HW3 by year end will be an amazing achievement. The cross country trip by any of us is another pipedream, not that I care.
Another scenario is if there is an accident on the highway and the lanes are blocked. Your car may be stuck behind a police car with flashing lights and an officer directing you to go around. A normal person would know how to handle this but the car would just stay behind the police car until it was clear to go.
Ok, so let’s not get locked into a literal non stop drive coast to coast...I will have to pull over to pee, so I guess that breaks the FSD journey too. ; )
Accidents, potholes, debris, sure, all these things are potential sideliners, although the FSD presentation seemed to indicate that these kinds of things would eventually get recognized and handled. This year? Perhaps not. Likely not.
The great thing about setting goals - even ones that seem unrealistic - is that it gives your team something to shoot for, and if you miss it, then you have benchmarks for why. And if you reach them, all the better!
Stop with all that positivity and rational thinking, m3phan. Haven’t you heard, it’s all impossible and if we’re lucky, FSD will be here in 10 years.
Forget driving on Earth. Tesla needs to make cars that are rocket ships that can land on pads after being shot hundreds of miles into space!
@82bert ; )
Vetted route with lead car making sure there are no unforeseen real life obstacles maybe. Long long way off for true level 5 driverless autonomy. Stock pumping is what it’s all about.
Howard to the rescue. See, m3phan. Your pipe dream ideals are just that. It’s all smoke and mirrors. The man who helped build a startup to a legit auto and tech company in a matter of a decade is a fraud. Ignore the fact that no one has built an auto company in the US in 100 years. Everything he’s done and continues to do is just an illusion. Don’t you see?! Ha.
Bert, your thigh it’s not mine. He has accomplished a great deal. Part of that is his incredible vision and drive. Just being realistic. It will get there but not this year or the next. Sure some version might be sooner but not level 5 driverless LA to NY.
Howard, your history is clear. The words and implications are all yours. No more needs to be said. Fortunately for many on this board, there is no simplistic way to see a person’s historical posts. Your history on here speaks volumes. Best of luck to you.
I’m seriously all in for the ev aspect and co-pilot assist but don’t get musk’s push to make it sound like autonomics is just around the corner. That is a far bigger thing than everything he has accomplished all together. I guess that’s why he keeps focused on it alone. The rest of the stuff is comparatively childs play.
For those that don't think level 5 is possible anytime soon, might be worth looking at this Cruise Automation video. Totally different hardware and software than Tesla, but shows what can be done today: https://www.youtube.com/watch?v=6tiyZXKwdOA
Deals with all sorts of nasty situations running around San Francisco, including triple parked cars, construction and more.
Funny to see the first double-parked car is a Model S too!
Not sure if they didn’t run over the pedestrian at the 1:20 mark. Im skeptical whenever the change the speed of the video.
@TeslaTap.com "For those that don't think level 5 is possible anytime soon, might be worth looking at this Cruise Automation video. Totally different hardware and software than Tesla, but shows what can be done today: https://www.youtube.com/watch?v=6tiyZXKwdOA"
Wow! I lived in San Francisco for several years and never saw traffic as bad as they showed in this video, at least not on one drive. They obviously took many different scenes and selected the worst ones to show off their capability. Makes me wonder which scenes did they leave out; how often did the car make a mistake or just give up?
Nevertheless, it's pretty amazing how well it does.
Bert, So you would move the front seats forward sit in the back all relaxed and from down town LA tell navigation to drive you to down town NY? You really think this is happening this year? You are that trusting if it is? We will assume it is fully legal to do so.
@howard. That is my plan. Get into my Robotaxi next year and go.