Summon seems like a silly party trick to me

[SandiaGrunt]

Hey, you know what? My toddler was (almost) knocked down by an ICE car being reversed out of a parking space in our communal parking lot at home.

This is whataboutism. The existence of some bad human drivers does not mean we should permit bad technologies on our roads.

Smart Summon is objectively worse than human drivers. It can't reliably detect curbs, trash cans, garage walls, grass, reversing vehicles, stop signs, parking space lines, or semi trucks.

The really nasty thing about Smart Summon is that each use actually removes a decent human driver and replaces them with a much, much worse parking lot Roomba, controlled by people who predictably abuse it by letting it go outside their line of sight.

Know that technology is never perfect and adjust accordingly.

This technology isn't just imperfect, it is objectively terrible. The appropriate "adjustment" should be action by NHTSA/NTSB to universally disable it.

 

[MountainRatMat]

Parking lots are inarguably full of children.

I'll bite and argue the point...I passed about 100 parking lots on my way to work today and only one was full of children, actually I take that back, only the drop-off zone was full of children, the parking lot had no children in it. In addition, when there are children in a parking lot they are almost always accompanied by an adult. At 5mph a parent can pick up their child and run if this so called zombie Tesla you speak of turns into a T1000 and actively hunts down children without stopping.

Please don't let your 1 year old that walks better than a 2 year old roam in parking lots alone.

Sorry in advance for the snark, I couldn't help myself.

 

[Kant.Ing]

Yesterday I ran past a summoning 3 knowing it would stop, so the car stopped and I heard a human yell from somewhere. It seemed to me I have hurt her child in a parking lot.

 

[SandiaGrunt]

In addition, when there are children in a parking lot they are almost always accompanied by an adult.

"Almost always," except when they aren't. Post #37 in this very thread is a story from a fellow who let his toddler get away from him. It happens often.

It is not acceptable to tell parents everywhere that they must now clutch their children even more tightly... because you want to play with a shitty, half-baked, nearly useless feature that you know is objectively worse than human drivers.

The jury isn't going to buy it.

 

[diplomat33]

We've seen this movie before folks. It's the cycle of life when it comes to OTA updates. Tesla releases a brand new feature and some owners blast it as shitty and half-baked. Over the next few months, Tesla collects data from the fleet and incrementally improves the feature over the next few OTA updates. Some Tesla owners start to warm up to it. It's better but not good enough. Then, eventually the feature gets really good. We saw this pattern with AP2 when it was first released, with NOA and we will see this pattern with Smart Summon too.

 

[SandiaGrunt]

We saw this pattern with AP2 when it was first released, with NOA and we will see this pattern with Smart Summon too

But Smart Summon is fundamentally different than Autopilot.

I'm not really too troubled with incrementalism in L2 ADAS development, like Autopilot. There's a driver in the seat, nominally focused on the driving task, with good visibility and a variety of choices in handling disengagement. The driver is guided by their own self-preservation instincts to remain vigilant.

Smart Summon completely removes the driver's visibility, and replaces the space of possible actions with, literally, a single button. It removes their self-preservation instincts, too.

The predictable abuse of Smart Summon is that people will let it go completely out of their sight, and that has already happened multiple times. When the car is out of their sight, they are no longer "controlling" it in any meaningful way. We know that Smart Summon is objectively worse than human drivers, too.

It is not acceptable to hand-wave these fundamental differences away with a line like "but it will get better eventually."

 

[diplomat33]

But Smart Summon is fundamentally different than Autopilot.

I'm not really too troubled with incrementalism in L2 ADAS development, like Autopilot. There's a driver in the seat, nominally focused on the driving task, with good visibility and a variety of choices in handling disengagement. The driver is guided by their own self-preservation instincts to remain vigilant.

Smart Summon completely removes the driver's visibility, and replaces the space of possible actions with, literally, a single button. It removes their self-preservation instincts, too.

The predictable abuse of Smart Summon is that people will let it go completely out of their sight, and that has already happened multiple times. When the car is out of their sight, they are no longer "controlling" it in any meaningful way. We know that Smart Summon is objectively worse than human drivers, too.

It is not acceptable to hand-wave these fundamental differences away with a line like "but it will get better eventually."

That's a valid concern now. But it will be less of a valid concern as Smart Summon gets better. For example, when Smart Summon does become as good as a human driver and can be trusted out of line of sight, then your concern won't be an issue any longer.

 

[SandiaGrunt]

That's a valid concern now.

I’m glad we agree. Get this **** off the road for now.

 

[MountainRatMat]

"Almost always," except when they aren't. Post #37 in this very thread is a story from a fellow who let his toddler get away from him. It happens often.

It is not acceptable to tell parents everywhere that they must now clutch their children even more tightly... because you want to play with a shitty, half-baked, nearly useless feature that you know is objectively worse than human drivers.

The jury isn't going to buy it.

The author of that post admits it was their fault the child was unattended AND this wasn't even a Tesla that almost ran over the child. I feel you are conflating two different issues that aren't even related to support your argument.

 

[SandiaGrunt]

admits it was their fault

It doesn't really matter who accepts fault -- it will still be a tragedy when a toddler is eventually killed by Smart Summon.

wasn't even a Tesla that almost ran over the child

Right, and Smart Summon is objectively worse than human drivers. This should concern you.

 

[alsetym]

I’m glad we agree. Get this **** off the road for now.

Yep, lets all go hide in a nice safe cave! The US literally allows people to walk around with guns strapped to their side, they can misuse them anytime they want to kill someone! Heck, ladders kill more people than Teslas in parking lots. Ban Ladders!

As far as I've seen a Tesla hasn't actually hit anyone yet (someone else backing into one doesn't count)...so sounds like you're getting your knickers in a knot for a theoretical danger where as what we have seen is that unless you actively try to get run over, they are probably better at avoiding hitting a kid than many drivers.

The world is a dangerous place. Get over it.

You also don't have to be aggressive and rude. Standing behind a keyboard does not give you an excuse to forget civility.

 

[alsetym]

It doesn't really matter who accepts fault -- it will still be a tragedy when a toddler is eventually killed by Smart Summon.



Right, and Smart Summon is objectively worse than human drivers. This should concern you.

So, what happens when eventually many many fewer toddlers re run over in parking lots because of this technology? You keep saying that smart summon is objectively less safe. I challenge you to prove it. How many accidents are in parking lots every day compared to how many smart summon accidents? Don't confuse objectively less quick and clunky with less safe. Objectively means numbers...which you have none, so it is purely you subjective opinion. Which is not worth any more than any other opinion on here.

 

[SandiaGrunt]

I challenge you to prove it.

Sorry, it just doesn't work that way. Tesla doesn't get to unleash whatever they want on the world, and then force the world to prove that it's safe.

We are a society with laws. There are legal precedents. From NHTSA:

"Unreasonable risks due to predictable abuse or impractical recalibration requirements may constitute safety-related defects. See United States v. Gen. Motors Corp., 518 F.2d 420, 427 (D.C. Cir. 1975) (‘‘Wheels’’). Manufacturers have a continuing obligation to proactively identify and mitigate such safety risks."

 

[diplomat33]

it will still be a tragedy when a toddler is eventually killed by Smart Summon.

Yes, it would be a horrible tragedy if it happened but it has not happened. Sadly, 205 children, 14 years old and younger, were actually killed by cars that are not smart summoned according to a 2007 report. Source: Study: Parking lots present high risk of injury, death in children due to lack of attention

 

[SandiaGrunt]

Sadly, 205 children, 14 years old and younger, were actually killed by cars that are not smart summoned

This is whataboutism. The existence of bad human drivers does not mean we should allow bad technologies on our roads. The existence of some tragedies does not mean we should accept more tragedies.

The legal precedents on this topic are clear, @diplomat33.

 

[electronblue]

As far as I've seen a Tesla hasn't actually hit anyone yet (someone else backing into one doesn't count)...so sounds like you're getting your knickers in a knot for a theoretical danger where as what we have seen is that unless you actively try to get run over, they are probably better at avoiding hitting a kid than many drivers.

We already have Smart Summon backing into another, stopped car.

It was blocked / out of sight of the owner if I recall.

 

[electronblue]

Yes, it would be a horrible tragedy if it happened but it has not happened.

It is too soon to judge either way.

 

[electronblue]

The world is a dangerous place. Get over it.

You also don't have to be aggressive and rude.

Speaking of rude...

 

[Big Dog]

Agreed. I’m at a loss on why there needs to be such venom in this by some.

If one doesn’t want to be part of Beta, then absolutely no one is forcing its use upon anyone and obviously the car itself can disable the feature set, in addition to the owner/user independently electing to utilize it....or not AND (importantly) where and when.

If someone doesn’t yet feel comfortable or confident in using it in beginning Beta (generally), they can still enjoy the sea of other features in V10 that are objectively...individually...and jointly pretty great and just wait out that refinement..then turn it on down the line.

That said, whether other V10 features like Netflix...YouTube...Hulu...or even going a minute back...Sentry (for example)...those are just...objectively awesome features no one else has in a vehicle and that for most, didn’t even exist when they purchased their vehicles! (but now are car features they’re able to enjoy daily and for the low monthly cost of...“free!”)

For me...-that- is value I appreciate and where FSD/Auto Pilot/Smart Summon are concerned, I enjoy being part of the feedback for Tesla that is going to ultimately transform the way people move.

Like Auto Pilot and Navigate on Auto Pilot, there’re works in progress (this...admittedly more consistently and visibly than the others presently) but one CAN simply manually stop the movement by just taking your finger off the App IF safety is the concern, as it’s supposed to be constantly monitored and in sight when in use anyway.
My experience (thus far) is that this part of the feature in Smart Summon actually works virtually instantaneously and well.

It WOULD be Great if it was better “ready for prime time”..sure...I think everyone agrees that would be wonderful...but that’s completely different than to bestow it essentially zero value when no one else has accomplished anything remotely close (in this AND in all Tesla has done and is doing with FSD).

Miles to go (yes) but for me, that doesn’t mean the miles travelled are valueless.

Rather, they’re foundation...and I appreciate being able to be part of testing and help refine, even though it still needs a lot of work.

I too, appreciate being part of the journey, and its fun to watch the baby steps and continual improvement and growth. That said, I find zero features of V10 that I will ever use. (well perhaps Joe Mode, but that's it). Can't wait for v.11. (back to the closet to don my fire suit for all the forthcoming Disagrees....)

 

[electronblue]

@Big Dog

I just disagreed to keep you warm and toasty!