Self-Driving Tesla Crashes into Wall Painted to Look Like a Road… Just Months Before Planned Robotaxi Launch

I'm trying to find an article that covers what I remember but I know for sure that it's been a good while since I saw the info I recall. Hopefully I can dig something up.

Indeed there is a lidar car in the video and it works way better in many scenarios.

Funny thing is, the price of lidar is dropping like a stone; they are projected to be sub-$200 per unit soon. The technical consensus seems to be settling in on 2 or 3 lidars per car plus optical sensors, and Chinese EV brands are starting to provide self driving in baseline models, with lidars as part of the standard package.

Cameras and radar, I believe. Never lidar.

Ah okay. I was genuinely curious if I was remembering correctly because I definitely know it's been awhile since I'd read anything on the subject.

Did he want to cut costs or did he want a network of cameras at his control all over the world?

meanwhile our subaru has lidar for adaptive cruise control and emergency braking

Yikes.

they’ll just physically unplug them the next time you have the car serviced

So, (looks at watch), in an hour?

I believe he claimed that since humans use their vision to drive that computer vision was more than enough.

I don’t know about you, but I also rely on sounds & feel when I drive. I also know that the human eye has evolved to detect motion, filter out extraneous information, and send just the important bits to the brain so that it doesn’t get overloaded with everything the eye sees. Computer vision is the exact opposite from that, having to process every bit of every image the camera sees.

Yep! That's what I'm thinking of. It was Elmo. The real engineers objected.

As a space nut I seriously hope that he never gets a chance to do anything similar with SpaceX. Thankfully he's mostly been kept away from important things thus far.

Don't get me wrong, I know SpaceX's closet is overflowing with skeletons. But since Congress has been so kind as to continuously cut NASA's budget for the last few decades, I have to rely on SpaceX and other private companies to keep our space endeavors going.

I did low-key get the squiggles before writing the article. I thought, from an ethical hacking disclosure-type perspective, that this info might cause folks to... well, ya know, paint tunnels on walls.

Then I looked, the cat was already out of the bag, the video had something like 5 million views on it in the 4 hours it took me to draft the article. So I shared it, but I definitely did have that thought cross my mind. I am also a little worried on that score.

It was removed due to supply chain, but Musk did seem to legitimately think optical only was better.

still, this should be something the car ought to take into account. What if there's a glass in the way?

A camera will show it as being more convincing than it is. It would be way more obvious in real life when seen with two eyes. These kinds of murals are only convincing from one specific point.

As much as i want to hate on tesla, seeing this, it hardly seems like a fair test.

From the perspective of the car, it's almost perfectly lined up with the background. it's a very realistic painting, and any AI that is trained on image data would obviously struggle with this. AI doesn't have that human component that allows us to infer information based on context. We can see the boarders and know that they dont fit. They shouldn't be there, so even if the painting is perfectly lines up and looks photo realistic, we can know something is up because its got edges and a frame holding it up.

This test, in the context of the title of this article, relies on a fairly dumb pretense that:

1. Computers think like humans
2. This is a realistic situation that a human driver would find themselves in (or that realistic paintings of very specific roads exist in nature)
3. There is no chance this could be trained out of them. (If it mattered enough to do so)

This doesnt just affect teslas. This affects any car that uses AI assistance for driving.

Having said all that.... fuck elon musk and fuck his stupid cars.

I mean, it is also referenced in the article and even in the summary from OP.

Or when a truck is moving traffic lights

https://youtu.be/QhA2CH6Z-v4?t=99

That's almost as bad as Sidewinder missiles locking onto the sun.

There's a lot of cars that check via camera too to double check, for missing/outdated information and for temporary speed limit signs.

Yikes, there’s a 25 around here that shows up as a 55 in Google Maps.

Also a 55 that goes down to I think 35 for just a moment when it joins up with a side road. I wonder what a Tesla would do if it was following that data.

Yeah, that is just how youtube works. You as an individual can say you don't like annoying thumbnails and titles, but they 100% work. And channels that don't use them are just not getting as many viewers.

algorithm-driven platform

And what is this "algorithm" based on? Actual user behavior. So the way to correct an algorithm is to change actual user behavior, no?

Still supports a creator pulling clickbait.
The only way is to vote with views/retention.

I don’t see a problem with thumbnails that accurately portray the contents of the video, since only a small number of characters can fit in the title and a screenshot of one frame from the video doesn’t say much, so it can be difficult to get a sense for the video at a glance otherwise. I do get really annoyed with thumbnails that are deceptive in any way. If the thumbnail seems like it might be deceptive, I’ll usually read the comments before watching the video, or quickly scroll through it to see if it’s BS or not. Sometimes, the thumbnail advertises something that happens at the end of a 20 minute video that could’ve been 30s, in which case, I’ll scroll usually through to the end instead of watching the whole thing. If it weren’t for the thumbnail, though, I might not have watched it all.

yeah but if you share it with people, they'll still see the clickbait thumbnail, and that's the actual problem

Thanks no I hadn't. Is that available as a Firefox extension. I do most of my browsing on desktop.

IMO it doesn't need to be a rigorous scientific test, it's not trying to prove something works as it should under all conditions. It's showing the exact opposite, it does not work under this one condition, which is more than enough to disprove the safety of the car.

Why would children be interested in car safety?

I know, but if they are about anything serious like tests, I think it's a fair assumption that the thumbnail represent it reasonably.
If it's misleading, I don't want their vomit. They can just fuck right off. We already have more than enough misinformation. I simply don't want to waste my time on bullshit.

History turned off, subscriptions only for me.

We are in a tiny, tiny minority.

That would require signing in and allowing tracking.

True, but Mark's video basically about comparing Tesla's Camera Sensors Vs Self Driving car with a Lidar Sensor.

They also simulated some real life scenarios which the car with Lidar sensors passed easily, while Tesla failed some of them.

So I guess Lidar sensors are superior compared to Teslas cameras.

Where on earth would you find a wall on a road with a fotorealistic continuation of the road printed on it?

Spoken like a man who has never relentlessly pursued a roadrunner, nor taken a wrong turn at Albuquerque.

You haven't seen what Teslas are in the news for lately?

It's not that crazy someone would put up a fake wall on some backroad to catch out inattentive Tesla drivers. Doesn't even need to be nearly as big and elaborate as this one.

But the point of the video is that optical cameras are easily deceived, and Elon is lying to his customers that LiDAR is overrated and not necessary.

This YT channel definitely went all out on the cartoonish nature of this particular test, but the article describes other tests as well including running over mannequins representing children that other cars (Lexus) avoided.

While I agree that this would trick many human drivers, I think the goal of a self-driving car is that it be better than human drivers. And there is existing tech that could help achieve that.

I actually agree, it's not really a good test. That wall is very realistic. It's just that people get pissed about negativity.

If it's made to be misleading and baiting, yes I FUCKING should. And so should you and everybody else.

Maybe, but it's not because of me people get shitty links.

Thanks. 😎
Then imagine why 15% downvote? I suppose it means they don't see how it's misleading?

But being dishonest about it, detract rather than add to it.

He did say lidar was "useless" though.

Tesla had camera+radar+sonar, and that wasn't their own tech - they used mobileye EyeQ back then. When they switched to in house tech they gradually ditched the radar and sonar which made no sense to me. But at the time I saw their lead say in an interview that this is superior and I believed. not anymore.

they said doing so cut costs but obviously lidar/radar/sonar only gets cheaper over time, let alone the extra r&d costs because a vision only system is much more difficult to develop.

It was removed because it was giving false positives. They should have upgraded it with lidar but decided to just remove it.

Except GPUs 😥

Leon said other sensors were unnecessary because human driving is all done through the sense of sight...proving that he has no idea how humans work either (despite purportedly being a human).

Yeah but the radar/lidar may allow them to drive into things quicker.

Wow you guys even lost the ability to do syntax. I guess it was only a matter of time.

I found this article. My takeaways were:

1. No driving in bus lanes during certain times of day.
2. No using the shoulder as a turn lane.
3. No using a bike lane as a turn lane.

I’ve said for a while that you could shut down an entire city with just a few buddies and like $200 in drywall screws. Have each friend drive on a different highway (or in a different direction on each highway) and sprinkle drywall screws as they go. Not just like a single dump, but a good consistent scatter so the entire highway is a minefield and takes hours to properly sweep up.

The thing is, we recognize a truck with stickers of a stopsign, while AI vision gets confused.

Lmao would it be illegal to put a stop sign on the back of your car?

Good.

I'm not rooting for America to fail, but I'm 100% rooting for Elon to fail.

Dude.

You could at least look at what you're replying to before jumping in and going full outrage mode. I didn't even say anything about what I thought of the validity of that experiment.

Keep putting yourself forward to defend the poor, misjudged car company belonging to a crazy asshole.

Easy

Username checks out.

It wasn't because of phantom braking.
It was to cut costs.

I don't know the value of echolocation in this case, as I'm generally ignorant here, but it's straight wild to me that they went purely on visuals.