Sure, technologically what they do have is pretty good. But considering how perfect those conditions were, and he is driving in grandpa mode, and still had to take over control of the car after just a couple of minutes isn’t totally inspiring if you’re talking about a car without a steering wheel and acceleration/brake pedals in 18 months to 3 years.
I’ll say it’s 3 years out once you have a successful winter highway test.
You mean perfect weather. But he is not driving in "grandpa mode" the car is driving.
If anyone is legit interested on what Tesla is doing watch this.
Anyway for people with an open mind and the technical knowledge to understand what is going on here, FSD is getting better and better.
It's hilarious how you frame every post you make around everyone else being an idiot, and only a savant like you could grasp what's going on. You know other people have access to the internet too, right?
Anyway, that video obviously had several issues. Is it pretty amazing? Hell ya it is. Is it ready yet? Absolutely not. I'd be interested to see how it handles unprotected lefts at highways speeds. It seems to not be great at judging speed/distance of approaching vehicles. And as Wormius mentioned, while a bit complicated, the conditions are perfect, with nice road lines everywhere.
I look at FSD the same way I look at SpaceX launching and landing rockets and reusing them. Everyone doubted it, but they kept at it till it worked.
There is something to be said for the most brilliant minds in the industry working on a super hard problem. It becomes a 'matter of time' problem, and not a '30 years away problem' like cold fusion is (strange comparison).
We already have forms of self driving, and we are already seeing better forms of it, including FSD, though not level 5. Do we really think as AI progresses that someone (maybe not Tesla) won't figure it out?
As for the whole 'other manufacturers will blow them out of the water argument', I think if it were really that easy, with the demand that Tesla has, other manufacturers would already be doing it. But it isn't, because it comes down to batteries and charging, and Tesla is far ahead on both fronts.
The argument that anyone can build a vehicle, especially companies who have been building vehicles for decades, is true of course. But that isn't what the issue is.
It's hilarious how you frame every post you make around everyone else being an idiot, and only a savant like you could grasp what's going on. You know other people have access to the internet too, right?
When people demonstrate at best surface level understanding of Tesla's FSD I point it out. I'm seeing that here, talking about weather and the like these things are very low down on the list of concerns. I already know what your reaction to this will be and further prove my point. If you think you know way more than me about FSD then great I'd love to hear all about it. No sarcasm.
Tesla is getting close to the point where what will hold back full autonomy is processing power and data, the structure of the system is largely complete. Have you listed to or watched any of the vids where the Tesla engineers talk about this stuff?
I was curious how it worked in winter, and found this one...pretty mild winter driving at slow speed, mostly 20mph.
Starts off steering right into a parked car @ 1:50
Randomly veers left into oncoming traffic lane @ 2:27
Turns left into oncoming traffic @ 3:45
wanders left and right @5:00
nervous moment @ 5:50
kind of a big fail @ 6:00
interesting it sees snow as "un-drivable area", this continues for awhile.
doesn't stop for stop sign @8:30
driving way to the left @ 9
wrong lane @ 9:50
I was curious how it worked in winter, and found this one...pretty mild winter driving at slow speed, mostly 20mph.
Starts off steering right into a parked car @ 1:50
Randomly veers left into oncoming traffic lane @ 2:27
Turns left into oncoming traffic @ 3:45
wanders left and right @5:00
nervous moment @ 5:50
kind of a big fail @ 6:00
interesting it sees snow as "un-drivable area", this continues for awhile.
doesn't stop for stop sign @8:30
driving way to the left @ 9
wrong lane @ 9:50
Sounds better than 95% of the driving I see in Calgary, LOL.
The Following User Says Thank You to sketchyt For This Useful Post:
Sounds better than 95% of the driving I see in Calgary, LOL.
Ya, a lot of it I was thinking, this is like distracted cell phone driving. But as an autonomous vehicle, I'd be a terrified passenger watching that go on.
A lot of weird stuff going on there. Clearly they had to defeat the autopilot safety features, so I don't really think this is to much of Tesla's fault. If they had a "driver monitoring camera" these idiots would have put a mannequin in the seat to get around it. Stupid people did stupid thing, won stupid prize.
Co-pilot makes a lot more sense than auto pilot. I am kinda curious how they got it up to such a high speed while not being behind the wheel. Dummies probably didn't even turn autopilot on, just hoped in and tossed a brick on the pedal. Wooo! look at it go!
Co-pilot is the perfect name Tesla dropped the ball here.
I predict it will come out this car didn't have full self driving and was not in autopilot during the crash. And there was a person in the drivers seat or possibly there was a sack of something on the seat and the belt was buckled.
I predict it will come out this car didn't have full self driving and was not in autopilot during the crash. And there was a person in the drivers seat or possibly there was a sack of something on the seat and the belt was buckled.
I actually meant to say the car did not have full self driving enabled at the time of the accident, not that it did or did not have the option. But I guess same result.
An interesting read on the state of AI systems and self driving cars.
Quote:
Recently a driverless shuttle was involved in a crash because it could not understand the intent of a human driver of a tractor trailer ahead of it who very slowly backed up for more maneuvering room, expecting the shuttle to also back up (NTSB 2019). The tractor trailer driver did not know the shuttle had no driver (nor did it have the ability to operate in reverse).
The driver had an expectation built over years of experience that the other vehicle would give way and be able to reverse, but the shuttle had no rule set to reference. This scenario seems simple for human drivers who understand the need to negotiate to resolve uncertainty but such abstract principles and the development of alternative action plans, even simple ones, is outside of the realm of ML-enabled systems, at least for the foreseeable future.
Such ambiguous situations happen regularly in the driving domain and often with much more dramatic and deadly consequences. There have been several incidents where Tesla drivers have been killed while driving on Autopilot, an automated driving assist feature, which failed to see objects directly in cars’ paths, and a pedestrian has been killed by an Uber self-driving car while undergoing human-supervised testing (Crowe 2016, Griggs and Wakabayashi 2018, Lohr 2016). In all these cases, the skill-based reasoning automated systems that relied on bottom-up processing failed, and deaths occurred because the inattentive drivers did not realize these cars still needed their top-down reasoning and judgment.
FSD 9 has been released, this one has radar removed and relies entirely on vision. You can see in the above video at ~16:10, first where it looks confused, and second, where it doesn't see the concrete pylons at all, and almost hits them. Radar would have detected these and avoided them, but they appear invisible to the cameras.
It's also hilarious that these clowns are out testing it, and say "this is the perfect situation to test late at ngiht with drunk people all over not following rules" or something like that. No, that's how you run someone over. It's kind of amazing any Joe blow can beta test a clearly unfinished safety product late at night in unpredictable situations, and random bystanders are the test subjects, with no say in it themselves.
You can see in the above video at ~16:10, first where it looks confused, and second, where it doesn't see the concrete pylons at all, and almost hits them. Radar would have detected these and avoided them, but they appear invisible to the cameras.
Maybe, but that is only scratching the surface. Radar sometimes cannot tell the difference between a stationary and moving object (look up phantom braking) which creates a massive problem, do you believe radar, vision, lidar, combine all 3 and code to a consensus.
Quote:
It's kind of amazing any Joe blow can beta test a clearly unfinished safety product
Beta is in the hands of very few people not every "Joe blow".
If you are actually interesting in this stuff look up Andrej Karpathy.
On Saturday, Musk announced on Twitter plans to greatly expand the beta software tests. He revealed that “due to high levels of demand” to join the beta program, users will be able to sign up by using the in-car display to visit the “Service” section and then pressing the “Download Beta” option.
