Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
They must have wasted billions on this by now. What a waste, it was also unbelievably stupid to even think anyone would buy, let alone it would be legal to sell a driverless car without a steering wheel. Currently I believe the laws are a qualified driver must be at the wheel with full controls, in a car with any automation.

This isn’t a phone with no headphone jack, no sane person would consider it to be ‘courage’ to buy a car with no controls…
 
  • Like
Reactions: Algr and maxoakland
If we can't comprehend the behaviour of one vehicle we certainly can't extrapolate the behaviour of a complex distributed system of many vehicles. And we definitely won't be able to deterministically assign a liability when something does go wrong. So the idea is a non-starter already.

That's not even a worry though as it's not the limiting factor of the idea. Really it's down to two schools of thought on systems design. First line of thought is you must design a constrained system with no capabilities and add features to it. Second line of thought is you design an open system and remove defects. The latter is where FSD sits. The initial condition has a set of defects which is unknown and they are only found via discovery, which means you have to literally drive it into a tree before the inference network determines that it is defective behaviour. Same as when you spray water on a cat that is trying to piss on the carpet. The work around for this, which is what Tesla are doing, is to attempt to add constraints around it. The result of this, as you've seen is the things clamming up suddenly or attempting to drive into stuff suddenly as applying constraints to an unconstrained system results in non-determinism.

As you can see this entire thing is a **** show. Ignoring where we are now with meat sacks at the wheel, which is also a **** show, it's better to start with a constrained system. That's another essay but ignoring the FSD and electric car hype, if we want to survive the next hundred years FSD isn't going to have an ounce of relevance. Adjusting society around centralisation, lower car ownership and utilising automated public transport is where we should go. Rails are a pretty good constrained system...
I made no assertion about how to correctly do FSD, only that the metric of interest should be a comparison between human and machine performance at the macro level.
An existing system that began with unknown defects and gradually improved is well illustrated in commercial aviation, where we can look back at a number of crashes (particularly in the 70's and 80's) that led directly to major changes in a variety of areas that have continually improved safety and reduced rates of casualty per mile flown (e.g., CRM).
When starting with a constrained system (as you suggest), the problem remains that we don't know what we don't know. Ultimately, we will learn from experience and address those unknowns.
 
"It will allow customers to do things like watch movies or multitask while on highways."

This is why CarPlay is on borrowed time. Car passengers are a captive audience and can be monetized. Most vehicle manufacturers will want to own their entire in-car stack. Why let Apple or Google make $ off your vehicles' passengers when you can. Eventually nearly all car manufacturers will be like Tesla with a custom UI ripe for monetizing. Location based ads, app & game stores, movie purchases/rentals, etc.
 
It's time for everyone old enough to read this sentence to accept that there will not be individual autonomous driving vehicles in their lifetime.
 
  • Like
Reactions: turbineseaplane
If Apple actually goes through with this it will be a big fail. They are Not a car company. They need to stick to small electronics.
In the past I posted that Apple was interested only in automotive sub-systems that could be sold to any car manufacturer. A lot us always doubted it was feasible for Apple to align themselves with a car manufacturer to jump into the business.
 
  • Like
Reactions: Tagbert
You know that's getting tired, right? Apple Maps navigation is exceptional in our area (BC, Canada). The only gripe that I have is it continues to insist on U-turns at intersections, which is not legal in our area.
Nope it’s not tired. Apple Maps has improved leaps and bounds and even has some cool unique features and incredibly detailed street views and iconic building look arounds etc, however even in a major city like london I have to default to google maps to make sure I’m getting the shortest most sensible routes and general reliable information. It has completely been unreliable in its route mapping on too many occasions. When you’re out casually it’s fine that you arrive late somewhere, but when business is involved, Apple Maps cannot be relied upon in my experience.
 
Focus on CarPlay. It's that simple.

Is anyone out there seriously waiting and wanting this unannounced Apple Car? Price point: $99,999.00

Perhaps they could do Carplay as a separate project, especially as the vast majority of the car engineers will have compentencies in other areas. Oh wait they are, it's almost as if this massive company don't need to Focus entirely on one thing at a time!
 
so, basically, apple will start making a car, that will be in the betweens of 70 to 90 thousands dollars, only to be categorized as obsolete by themselves every 5 years? that, in apple language, means no support, no parts.

Cook can socks my cook... I drive euro, btu with that premise, any honda civic from ANY year since it started production, will be a better option. I honestly would buy me 6 KIAs before buying one apple car... at least i know i can use those KIAs till i actually kill them. And that my friend, is what a real car is. I can synthetize the fact that they make technology crap to last only a few years due planned obsolesce, but there is no way i buy the same philosophy for a car. NO, NO AND NO.

I know some of you might get one... but once you do, remember this i say. Swallow your pride, take a deep look in the mirror, and contemplate the idiot that is looking at you. Specially if you went for the first gen :p hahahaahaha
 
Disagree. I think it will happen much sooner than you realize. Waymo and Cruise are basically already self-driving within geofence, and the advancement made by Tesla in that area is improving by leaps and bounds with every update. Full self-driving is definitely coming.
Definitely coming doesn’t mean it’s coming any time soon. Elon’s estimates keep missing and missing despite his claim that it’s basically a solved problem. No it’s not.
 
You are assuming Apple plans to build a complete EV from scratch. Much much much more likely they will partner with someone already making an EV and put their chips / sensors into it, slap an iPad on the dash and call it a day.

Do you really think Elon Musk visited Apple JUST to discuss twitter? LOL. and just days after this new announcement?

Elon has overspent and needs someone that understands supply chain to step in and make Tesla profitable.
I’m not an Elon fan, but if you think Tesla doesn’t understand supply chain you are grossly misinformed.
 
I’ll take a cybertruck with its breakthroughs in braking, steering, design, headlights, manufacturing, and bulletproof body and windows. Oh wait, wrong company.
 
Apple can’t even get Handoff to work properly for me 90% of the time and Siri thinks me saying “Change Pods” means “Play Dance Music in your Library.”

It will allow customers to do things like watch movies or multitask while on highways.

Yeah, nah.
 
There will be no cars that will be Full Self-Driving in our time life frame
I gather the issue (in general) is not the sensors, or the algorithms, or the computing power, but the energy needed to drive all of the electronics and processors. However, chips are becoming more efficient every day so I would imagine we will see this in our lifetime, at least for the rich and possibly for public transport. The systems do not have to be perfect - just better than people.
 
Apple wanted to create a self-driving car without a steering wheel or pedals, but has decided that such a plan is not feasible at the current time.

No wayyyyyyyyy!

Not technically possible? Didn’t see that coming. Just one misleading adventure that laughingly suckered in people.
 
So many tired jokes...

This decision is not set in stone. Their AI team could have a breakthrough next week, putting self-driving capabilities back on the roadmap. That's the cool thing with ML/AI, a world-changing tech really can be one weird trick away (I work in this space, it happens). Of course, ML can also be massively flaky, hence the legal nightmare.

Apple is pretty good about accessibility, you will likely be able to switch the control mechanism to something else — handicap support has been around for a long time.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.