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

How often did you use the upgraded AP functions

Do you think it’s worth the $7900 (CAD) upgrade

  • Yes

  • No

  • I’m thinking too


Results are only viewable after voting.
This site may earn commission on affiliate links.
So you would rather be enjoying driving in THIS than chilling out?:eek::rolleyes: Also keep in mind that this enjoyment comes at the COST of the lives of almost 40,000 people a year in the US alone. If everyone switched to FSD today that figure would probably be cut by at least ½ and would be very low in just a few years.


View attachment 425913

In that trafffic I would be miserable regardless if I was the driver or a passenger. I would more than likely be streaming the Comedy station to relieve stress. :D I do find that one pedal driving and brake hold reduce a lot of fatigue in the stop and go situations like pictured. Also, AP/ Low Speed follow Cruise Control does a pretty good job in this situation, a little jerky at times as it tends to overreact but still pretty good IMO.

If FSD will indeed saves lives I am for it 100% but I am still very skeptical at this point. Driving is a learned skill so I worry that all of these drivers assistants could actually create bad habits and increase risk for all of us in the short term. For example, you have a relatively new driver who is used to a FSD vehicle and then all of a sudden they need to drive a rental car for a week. It is like riding a bicycle with the training wheels on all the time and then you are handed the keys to a motorcycle.
 
  • Like
Reactions: Msjulie
If FSD will indeed saves lives I am for it 100% but I am still very skeptical at this point. Driving is a learned skill so I worry that all of these drivers assistants could actually create bad habits and increase risk for all of us in the short term. For example, you have a relatively new driver who is used to a FSD vehicle and then all of a sudden they need to drive a rental car for a week. It is like riding a bicycle with the training wheels on all the time and then you are handed the keys to a motorcycle.

I think we'll be facing issues similar to what is happening in aviation. As automation handles more and more, pilots start to rely on it too much letting their hand flying skills to deteriorate and when the automation fails they struggle to handle it because that's when you really need to bring your A-game.

The same with FSD or even with the current TACC/Autosteer. I wouldn't recommend new drivers to use AP features until they are very comfortable driving manually, and even then don't use it all the time, only on long commutes to reduce the workload. And stay alert.

True FSD needs to be extremely reliable to graduate from driver assist to full autonomy. I don't think we are close yet. There are some situations where even the human driver gets confused.
 
Last edited:
True FSD needs to be extremely reliable to graduate from driver assist to full autonomy. I don't think we are close yet. There are some situations where even the human driver gets confused.

As an extreme example, I was driving at night on I-5 a year ago over Tejon pass. There was a construction on the bridge where they were doing something in the middle and they split the lanes in two, but neglected to erase old lane markings and just painted new ones on top. No constructions cones either. It was dark and confusing, so I followed the ones that I thought made sense until I realized that I chose wrong when I saw some shadow ahead of me. I immediately merged to the right and I was glad I did because it turned out to be an unmarked guard rail. No reflectors, no lights, nothing (they fixed it next time I was driving there). Not sure how FSD would handle that.
 
As an extreme example, I was driving at night on I-5 a year ago over Tejon pass. There was a construction on the bridge where they were doing something in the middle and they split the lanes in two, but neglected to erase old lane markings and just painted new ones on top. No constructions cones either. It was dark and confusing, so I followed the ones that I thought made sense until I realized that I chose wrong when I saw some shadow ahead of me. I immediately merged to the right and I was glad I did because it turned out to be an unmarked guard rail. No reflectors, no lights, nothing (they fixed it next time I was driving there). Not sure how FSD would handle that.

Precisely, that is why I think FSD is not realistic unless all vehicles were attached to a track based grid or something. There are so many corner cases where the driver has to constantly supervise or take control if this or that happens. It is more realistic to think that these drivers aids will continue to improve and automate a lot of mundane tasks and provide additional "eyes" on the road but we will be required by law to sit in that drivers seat to make sure somebody is held accountable for what happens next.

Rear backup cameras are a good example of how useful technology gets mandated and becomes standard equipment in vehicles. As FSD features mature I would expect the same. At some point this tech will not be a separate charge but built right into the price of the vehicle as standard safety equipment. Most new cars now have Lane Departure Warning, Rear Cross Traffic Alert (not Tesla for some crazy reason), Blind Spot Monitoring, Backup Camera, Pedestrian detection, Dynamic Cruise Control, etc. Honda was charging for it, then Toyota made it standard now Honda is doing the same. Just a matter of time.
 
I have only driven my Model 3 with the FSD package for a couple of weeks, but I think this might be helpful to those considering it because I still remember what it was like to not have any features.

At the moment, its a different way of driving. Conceptually, its the same as cruise control. With cruise control, you know that at some point you have to disable it, because, well, I don't believe there is a cruise control which could handle an entire trip, including stop signs and turns.

Autopilot, and Navigate on Autopilot, are like cruise control in the same sense, that only far, far more advanced.

On a highway trip, autopilot is likely to be fully sufficient unless there is some sort of construction which results in the actual lanes going from well marked to not well marked or not marked at all. That is where it is as of today. Lane changes seem pretty advanced in terms of waiting for an opening and then making the change. Keeping distance from the car in front is just as good as a human would do.

The only thing it does not do well at the moment is shown on many Youtube videos, where the lane all of a sudden gets wider (like an on or off ramp, and you know its not "really wider" but the car thinks it is so it drifts into the center of the new, wider lane. Not an unsafe move but obviously incorrect.

Autopilot also handles well marked streets at the moment, even though it is not "meant" to be used on them. In my area there is street with about three lanes in each direction, all stop lights, for a good 10 miles. Forty to 45 speed limit. If you are behind another car, it handles this perfectly, slowing when the other car slows for a light, all the way to a stop, then speeding up away from the light.

I have not yet seen what it does when it would approach a stoplight with no car to key off of, it will probably either shut itself off, or, obviously, I would shut it off.

The Navigate on Autopilot "knows" when it leaves a highway. It turns of and goes back to autopilot based on GPS.

I am happy to have purchased it, because for my $6k I get to be part of the testing group to experience all the incremental improvements along the way.

But, it also makes perfect sense to wait until the entire FSD is functional and buy it then. I would not call this a "counter argument" but its the same with any new technology, at some point you know if you wait a year you will get either the same thing for less or a better product. But in this case there is actual fun in using it now, plus, unlike, say, a computer, this product updates itself.

I don't think its going to need any other sensors, and, if they figure out they need a camera up front I imagine they will simply add it. Cameras are not expensive and all you need to do is have a small hole in the bumper.

My advice? If you think the $6K or $7K really matters, don't get it. Especially if you are focused, like some posters, and what it does not do at the moment. I would not add it to a leased car.
 
  • Like
Reactions: Joey_82
I have only driven my Model 3 with the FSD package for a couple of weeks, but I think this might be helpful to those considering it because I still remember what it was like to not have any features.

At the moment, its a different way of driving. Conceptually, its the same as cruise control. With cruise control, you know that at some point you have to disable it, because, well, I don't believe there is a cruise control which could handle an entire trip, including stop signs and turns.

Well, most of the features you mentioned (Autosteer, Traffic Aware Cruise Control) are part of standard Autopilot package anyway. The question is whether lane change assist / auto lane change and NoA are worth extra 6K (well, plus Summon and Autopark but they are not very useful at the moment)...
 
  • Like
Reactions: raptor5244
Well, most of the features you mentioned (Autosteer, Traffic Aware Cruise Control) are part of standard Autopilot package anyway. The question is whether lane change assist / auto lane change and NoA are worth extra 6K (well, plus Summon and Autopark but they are not very useful at the moment)...

I couldn't really argue that those features are worth it now. What makes the decision interesting it (1) the features will expand, and (2) at some point, it will be much closer to FSD. And we don't know exactly how the features will expand, or when.
 
I was going to write a lot about how awful FSD has proven to be to date, but instead I’ll just say NO, it is NOT worth it right now, and worse, it may not be worth it during the lifetime of the car for which I (stupidly) paid for the option.

Autopilot - great...as long as you’re in rush hour traffic for a long period of time with no need to change lanes, and you’re never in the merging lane.

NoA - great...as long as you don’t mind being pinged to change lanes randomly and errantly on a constant basis. Then, once you cancel the bullshit suggestions, you’re also happy to instanteously receive those suggestions again, unless god forbid you have auto lane change turned on, in which case you will actually look like a drunk driver to others on the road.

Auto park - if you can even get it to activate, you should play the lottery. And even then, my 102 year old, totally blind grandmother can park a car faster than this sh**

Auto lane change — WORST. FEATURE. PERIOD.

Auto lane drift prevention - awesome if you love to be forced to use your turn signal literally every single time you change lanes to avoid being nudged and nagged by your car constantly.


And on and on and on.
 
Last edited:
The more I use AutoPilot and Navigate on AutoPilot, the more I learn how it drives and have become comfortable with it:
  • one lane into two - there is a brief, 1-2 seconds of vacillation and then it chooses a lane. This is just the short-distance, vision of AutoPilot making a last minute choice. I'm no longer surprised as I know it will choose in a second or so.
  • poor lane markers - it won't go into AutoPilot mode and provides an unambiguous alert. Going into TACC first and waiting a couple of seconds and often improves the probability the car will go into AutoPilot.
  • curve away right lane marker or curb - still a problem but easily recognized and manual override works.
Some are unhappy AutoPilot does not drive the way they want. For them, manual driving works perfectly fine just like all the other cars on the road. AutoPilot does not meet their expectations and that is not uncommon with any product. In contrast, I prefer to learn how AutoPilot works.

I paid $3k for basic AutoPilot and expected there would be a 'learning curve.' With practice, I'm becoming comfortable with it and recognize how it has improved my driving experience. It has reached a point where driving our backup car, a BMW i3-REx, I have to remind myself that it does not have auto-steer. The BMW 'magic eye', dynamic cruise control works like TACC but it won't stay in its lane. It shares many of the same limitations of AutoPilot.

In engineering we have a concept of 'managing user expectations.' Sometimes we know enough about the requirements that we can address them before and demonstrate them concurrent with deployment. Other times, we learn about undocumented requirements that could not have been learned without a 'hands on' site visit. AutoPilot has reenforces the importance of managing user expectations.

It would be great if someone who is not Tesla, say a game maker, made a version that has AutoPilot and Navigate on AutoPilot available ... as a training or pre-purchase tool. The next alternative is to stitch together a set of videos showing AutoPilot warts which is what I do. But then I use it to document regression tests needed after each update.

Sure enough, this YouTube just popped up:
I'm not sure his methodology gave useful insights but it does reflect owner attempts to map the system and performance. This is a good thing.

Bob Wilson
 
Last edited:
.....
Sure enough, this YouTube just popped up:
I'm not sure his methodology gave useful insights but it does reflect owner attempts to map the system and performance. This is a good thing.

Bob Wilson
Watched yesterday day but.....

  • All the MCU stuff like rebooting and playing games have nothing to do with the FSD chip. Am I wrong????
  • Also comparing an X with 15.x to a 3 with 20.x seems like apples and oranges.
  • Don't think the code is optimized much if any yet to take advantage of the 3.0 FSD????
 
Watched yesterday day but.....

  • All the MCU stuff like rebooting and playing games have nothing to do with the FSD chip. Am I wrong????
  • Also comparing an X with 15.x to a 3 with 20.x seems like apples and oranges.
  • Don't think the code is optimized much if any yet to take advantage of the 3.0 FSD????
I would be more impressed if he (or someone) had a video that measured how quickly objects are detected. For example, stopping at a rail road crossing and using a video to show how quickly the 'train' image shows up versus the concurrent video of the train passing across the front.

Bob Wilson
 
So you would rather be enjoying driving in THIS than chilling out?:eek::rolleyes: Also keep in mind that this enjoyment comes at the COST of the lives of almost 40,000 people a year in the US alone. If everyone switched to FSD today that figure would probably be cut by at least ½ and would be very low in just a few years.


View attachment 425913

Yeah, exactly. I don’t know why this bizarre FSD-shaming concept of “I actually, you know, enjoy driving my car!” seems to keep popping up on these (and other) forums. I can only imagine the people saying that are either 1. Teenagers who haven’t been driving long enough to be tired of driving yet, or 2. So rich that you have a chauffeur for all of your day-to-day commuting needs, and the only time you actually drive a car is when you take your McLaren out on the track.

Everyone else, especially those of us in Southern California, are not enjoying sitting in 2+ hrs of traffic nearly every day whenever we want to go virtually anywhere other than our neighbors house.
 
Yeah, exactly. I don’t know why this bizarre FSD-shaming concept of “I actually, you know, enjoy driving my car!” seems to keep popping up on these (and other) forums. I can only imagine the people saying that are either 1. Teenagers who haven’t been driving long enough to be tired of driving yet, or 2. So rich that you have a chauffeur for all of your day-to-day commuting needs, and the only time you actually drive a car is when you take your McLaren out on the track.

Everyone else, especially those of us in Southern California, are not enjoying sitting in 2+ hrs of traffic nearly every day whenever we want to go virtually anywhere other than our neighbors house.

Some people actually enjoy driving. I agree that sitting in stop and go traffic is not my idea of fun as a driver, but it sucks pretty bad as a passenger as well. You make it sound like sitting in an air conditioned car while streaming music or talking on the phone while having to manage two pedals and steer is torture. Maybe we all take bus or just go back to walking?

I don't think many are arguing against FSD and its future benefits. I enjoy driving but if I get tired then having a system take over would be fantastic. It is just that FSD needs to be equal or better than the human driver for it to be widely accepted, otherwise it is just adds more work as we have to supervise the system and drive the car.
 
  • Like
Reactions: Msjulie
OP - I’m also in Canada and thinking about the $7900 FSD option (though I haven’t ordered my Model 3 just yet).

While I have had time to try it out, I did like NOA, auto lane change, and auto park quite a lot. But I think I’m going to wait a bit to see what other changes come to it. It is nice that its the same price regardless (whether buying a new car or adding afterwards). Of course it can go up in price in the future, but I imagine we’ll have some notice before it does. Given that I’m going to keep the car for a decent amount of years, and that I’d like to use these features, I’d definitely buy FSD before a price change/increase.
 
OP - I’m also in Canada and thinking about the $7900 FSD option (though I haven’t ordered my Model 3 just yet).

While I have had time to try it out, I did like NOA, auto lane change, and auto park quite a lot. But I think I’m going to wait a bit to see what other changes come to it. It is nice that its the same price regardless (whether buying a new car or adding afterwards). Of course it can go up in price in the future, but I imagine we’ll have some notice before it does. Given that I’m going to keep the car for a decent amount of years, and that I’d like to use these features, I’d definitely buy FSD before a price change/increase.

Agreed, it also depends on how long it takes. We are on the bleeding/cutting edge right now for FSD. If it takes long enough you may want to put that $7900 toward a Tesla Model 4 or whatever the future holds. It takes both great software and great hardware to pull off FSD and both will continue to get better over time.
 
Agreed, it also depends on how long it takes. We are on the bleeding/cutting edge right now for FSD. If it takes long enough you may want to put that $7900 toward a Tesla Model 4 or whatever the future holds. It takes both great software and great hardware to pull off FSD and both will continue to get better over time.

That’s a great point as well. Where I am FSD would be ~$9000 including sales tax. If planning on switching the car a few years down the road that’s a significant chunk of $ that could be saved or used in the future.