TexLaw
Member
It's totally an option, and they pretty much let you know that no one knows what it will do or when it will do it. If you want to take a swing and pay less up front, you're welcome to do so, or you can wait and pay more.
-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Should FSDC (Full Self-Driving Capabilities) be an option right now?
- Thread starter CTgigawatt
- Start date
-
- Tags
- auto pilot 2.0 fsdc Model S
P100DHG.
Member
I bought FSD I had a 90D on order without FSD. Moved my reservation and bought a P100D with FSD. I'd be completely content with level 4 FSD. If I can sit in the car and be a passenger and it takes me to my destination without me touching a thing, that's freakin awesome. Level 5 is something I don't need or really want. If my P100D just drove it self around without me in the car what's the point? Seriously though not a rhetorical question, I'm open to advice here.
Electricfan
Active Member
May the force be with you. You are where I came extremely close to being. I hope very much to see FSD working ASAP, but truthfully I can't see it working before 2020, if then. Its just an extremely tough thing, and just the trouble they're having with "auto-steer", something mobil-eye did for them with AP1 but they can't duplicate on their own, demonstrates they will need a lot more time to get FSD done.
It's actually a very Obama like move. See, .e.g.,
"If you like the plan you have, you can keep it. If you like the doctor you have, you can keep your doctor, too. The only change you’ll see are falling costs as our reforms take hold."
TexLaw
Member
I agree that I don't really even want level 5. I still like driving. Sure, it would be nice from time to time, but I still just like driving. That's why I got this here car that's fun to drive!
Ah yes - a great specimen of one of the loveliest of all post types - the high horse morality post cross bred with the emotional validation seeking post. This post type goes as follows:
"I do not approve of a certain transaction between two parties. Therefore it is immoral. I want validation for my feelings - do you feel the same way? Let us revel in our disapproval together - it feels so good."
I fail to understand the point of offering the FSD package when they did.
Let's look at it from their perspective when they released it
You have a massive hurdle to overcome with migrating the existing AP capabilities to HW2
You already have the best semi-autonomous capable car that's available on the road with AP.
The hook for HW2 is EAP, and EAP by itself is pretty impressive.
There is very little that is known about what the requirements will be for full-self driving.
The accounting rules prohibit you from declaring any revenue from the FSD package until it's functional.
You're inevitably going to get sued if you don't deliver.
What's the upside? Some media attention?
The risk greatly outweighs the reward.
Let's look at it from their perspective when they released it
You have a massive hurdle to overcome with migrating the existing AP capabilities to HW2
You already have the best semi-autonomous capable car that's available on the road with AP.
The hook for HW2 is EAP, and EAP by itself is pretty impressive.
There is very little that is known about what the requirements will be for full-self driving.
The accounting rules prohibit you from declaring any revenue from the FSD package until it's functional.
You're inevitably going to get sued if you don't deliver.
What's the upside? Some media attention?
The risk greatly outweighs the reward.
No, it just demonstrates you don't remember how long it took to release AP 1.0 after 1.0 hardware began shipping in cars (a full year). You're also the guy who says AP 1.0 induces white knuckles. AP 2.0 is 5 months post hardware release and rapidly improving. I have AP 1.0 (no white knuckles here - just thousand of miles of relaxed driving) and an AP 2.0 arriving soon. If, at 12 months post-release (October 2017) my 2.0 car has not equaled my 1.0 car's current capabilities - I'll grant your point.
Also @Electricfan I imagine FSD will be released in small baby steps just like autopilot:
Stage 1: "Okay those of you who paid for FSD may now remove your hands from the wheel on divided highways at speeds up to 65 mph."
Stage 2: "Stage 1 + 10 mph"
Stage 3: "Undivided highways"
Stage 4: "City driving in pre-defined geographic areas on sunny days during the day" - ie major metros - SF/LA/NYC
etc etc etc - with stages of FSD showing up every 6 months or so. I for one will be delightedly happy if this is the progress we can expect and won't feel ripped off one bit.
Last edited:
So says the Tesla hobbyist on the internet. The man who built Tesla (you remember - the one who started building electric sedans when people told him it was a fools errand - and then launched autopilot to much hobbyist-hand-wringing) and who is landing rockets upright on barges says otherwise. If Musk had listened to reasoning like this he would never have gotten anywhere in life close to where he is now. YOU have no idea what the requirements will be for full self driving. Musk has considerably greater access to and influence on the power brokers of the world than you do - and I imagine he knows more about machine learning than you do as well. Oh - and if you think "enhanced auto pilot" is a sales hook that gets people excited, dreaming and keeps Tesla's name burnished and sales driving forward - don't go into marketing as a career.
Not attacking you but the history of Tesla is filled with this exact line of reasoning - it's at the point where it's funny now to hear it again. "I don't understand X therefore Tesla should not do Y because the downside is greater than the upside." Meanwhile the beat goes on and Tesla keeps growing...
Last edited:
You didn't really approach that the way I expected.
Put yourself in Tesla shoes at the time, and put forth your arguments FOR offering FSD.
Don't pull this Elon knows best crap.
I know a little something about machine vision, and for machine vision to work it has to see. It hasn't yet been demonstrated that the HW2 cameras work well in the rain. I'm still waiting for an HW2 owner to report how well the rear camera does in the rain.
Last edited:
I wouldn't have pulled the "Elon knows best" crap 5 years ago. But the guy has done nothing but win for almost a decade (if you define "nothing but win" as building Tesla into a continual growth machine that is beginning to look like a juggernaut that's accelerating away from the competition). At this stage in Tesla's development I am willing to give him the benefit of the doubt based solely on his track record and I don't think that's foolish reasoning.
To take your bet the track record is clear - a long history of people saying "I don't understand how this can work - therefore the CEO doesn't either and I conclude the risk outweighs the reward." Those people have a long track record of losing against Musk.
Now - as for the forum nitpickers who like to talk about the imperfections of AP 1 and the things he said about it that he did not deliver - sure, they are right on their nit picks. But they are totally irrelevant human beings to the story - and if you can't see that you're staring at your fellow forum nerd trees instead of the huge human forest that is the market at large. If Musk gets even SOME truly hands off capability in hardware 2 under SOME conditions - he'll win with the market. Nobody expects this suite to deliver foggy/snowy driving. But sunny day driving in well mapped areas? Yeah - I think he'll get there with this suite. He has ALREADY implied he designed it to have a swappable board - if more teraflops are needed, we'll get more teraflops.
To believe that you are correct @S4WRXTTCS one must believe that a pure 360 degree camera and forward radar system is incapable of statistically safe full self driving (ie death rate similar to or lower than human driving) - no matter how many teraflops are available and no matter how much training time has been put into it - even under sunny day conditions in geofenced areas of major metro markets.
Your scenario is implausible. As for regulations? That's different but look which way the wind is blowing. He's had huge victories already with the 40% AP 1 accident reduction rate - that's a PR win I really think folks here underestimate. That win gives regulators political cover to grant Musk the benefit of the doubt and at least give him reg approval under certain conditions on a trial basis.
Last edited:
Ulmo
Active Member
I'm trying to find the right time to ask for a new financing feature, and debating whether to hijack this thread for it. No, thread hijacking is not nice. I'll wait until near release. (btw, the idea is to allow Tesla Network uber-like self driving service to "rent to buy" the Full Self Driving feature, only renting it while it is driving Tesla Network revenue passengers even if an owner did not pay the $10K for the FSD option during regular non-Tesla Network use; it would be a % of the uber-like network take and built equity until the feature is fully bought, at which point it would be available to the owner for non-Tesla Network use as well, as well as no longer be a % of the take. I give permission for someone else to start their own thread for this. I'm thread starting shy right now.)
I see it as a $10K option, since that's what it costs after delivery right now without the prerequisite option, and all together it would be $10,000.00. I know you already had EAP, but since I don't even have that, I see it as including the price of both. For $10K, one could buy a used Leaf for that price that performs more functions, or a whole new car for that price (albeit not electric), or one hell of a lot of Uber & Lyft rides, and eventually, Tesla Network rides.
Last edited:
It hasn't been demonstrated to YOU that the cameras work well in the rain. In addition - I'll readily grant that this system may be sunny day only. The argument for offering FSD - okay I'll lay it out but I thought it was obvious:
1 - Huge PR win to offer self driving. HUGE. You can't buy publicity like this. Musk maintains the fearless leader persona and Tesla retains the mystique. Enhanced autopilot has no mystique, no sex appeal.
2 - Prepares for Model 3. This one is obvious isn't it? The S/X fleet serving as development mules prior to release of Model 3. First Model 3 deliveries will be roughly 1 year after HW 2 release into the wild. A full year of development and fleet learning before the unwashed masses take the wheel.
3 - It's the Musk way - make a big promise in public as a way of cracking the whip on yourself - then work til it kills you to fulfill it. What's the argument for landing rockets on barges? For building an electric sedan as a tiny start-up with little money at a time when nobody thought it was viable? For building a proprietary rapid charge network across the USA? For being first to market with AP 1.0 when everybody was wringing their hands about how risky it was "THINK OF THE CHILDREN!" Your argument of risk outweighing reward could have been applied to ANY of Tesla or Space X's major milestones. You've read his biography right? You know how the guy thinks, right? You think he's gonna change now?
4 - It's driving sales. I know because I'm buying one - despite having 1.0 already. This is a software problem - it will be solved. The hardware is in the cars - 360 degree cameras + a supercomputer.
Will it be a couple years before I get the notification I can remove hands under some conditions? Maybe (though I doubt it based on how rapidly the suite is improving already). I'm cool with that.
$10K is peanuts if you value your time at even a modest hourly rate, and then sum up even a fraction of that hourly value over the lifetime of the vehicle.
Ulmo
Active Member
I debate that. Drivers sit in the left (or right in UK) of their cars, and can see down the line of cars in front of them if they bob left (right in UK) a bit to see around the ones immediately blocking their view. Drivers can use this distance lookaround to gauge traffic in lanes up front for miles ahead, blocks ahead, and many cars ahead, for many lanes. The Tesla hardware is handicapped by being centered in the windshield and not in the sides of the car like the driver.
Yes, I agree. I'm just pointing out the option is a lot more expensive than the $3,000 quoted in OP's message in the circumstance of many buyers.
Ulmo
Active Member
I know they thought of it. We all get to see how it manifests in the future, obviously, and I can't claim to know the future, only guess.
To add to the arguments I posted above for offering self driving now:
5 - Learning - Tesla now has a first mover advantage in terms of fleet learning. By the time anybody else in the industry gets around to shipping a suite they say is capable of Level 4/5 in normal conditions and can fleet learn - Tesla will be on V3 of the Autopilot suite (perhaps containing Lidar) and will have two years or more of learning behind them - along with a fully proprietary vision system the fleet began testing way back in October 2017.
Finally @S4WRXTTCS @Electricfan @Ulmo and others - remember this suite is shipping in the Model 3 (perhaps + a HUD). Tesla is fully aware the 3 is the most important launch in their entire history. They know the stakes are huge for getting this right. Given that the financial stakes of NOT delivering at least a modestly capable full self driving experience that they have promised to hundreds of thousands of reservation holders (soon to be owners) are gigantic - it is reasonable to imagine that a number of highly qualified engineers at Tesla and Nvidia believe this approach will work. Again, if asked to put money down on a bet whose bet would you take? A few outside hobbyists on the internet with no inside access to Tesla's development team - or the bet of the company with a strong track record of delivering tech to market others thought impossible, and which has billions of dollars riding on the outcome?
I'll take Tesla's side of the bet.
I'll take Tesla's side of the bet.
Similar threads
