Scenarios in which Tesla FSD Beta 9.0 fails • Hacker News Board

Re: Scenarios in which Tesla FSD Beta 9.0 fails

#111

Post by mikestew » Mon, Jul 12, 2021, 6:41 PM UTC

darknavi wrote: post #62

H8crilA wrote: post #54

Earlier quoted context omitted.

You know there's a very old and working solution to this specific problem, it's calling the police and reporting the license plate numbers so that the driver is punished for reckless behavior. Technology made it much easier via dashcams.

I've admittedly only tried doing this once, but it got no where. The police said they wouldn't/couldn't do anything unless they were on-site and could ID the driver.

Same experience, which causes me to question the utility of signs along (as one example in the Seattle area) East Lake Sammamish Parkway stating that I should report aggressive driving. Well, not if you're just going to blow me off.

Re: Scenarios in which Tesla FSD Beta 9.0 fails

#112

Post by belter » Mon, Jul 12, 2021, 6:43 PM UTC

toomuchtodo wrote: post #81

belter wrote: post #75

Earlier quoted context omitted.

Show me Self driving on a closed track with snow or rain please.

Does open road snow count? https://youtu.be/fKiWM1KjIm4 https://torc.ai/torc-self-driving-car-dashes-through-snow/

Its an interesting video but its an edited video.

The company website mentions the system was able to negotiate the challenges of the driving conditions, looks like an euphemism...

Very few details, no scientific publications I could find on their Asimov system on a quick search, not sure if this is a technological breakthrough or a fine tuning of existing processes and methods.

Because if its a fine tuning of current algorithms not sure how long they can push this. They are relying on Lidar (and other sensors) but even as of last year, it seems most teams already realized Lidar would not be the solution for Snow, and were now pushing Ground Penetrating Radar ( sounds expensive...)

"Autonomous Cars Struggle in Snow, but MIT Has a Solution for That" https://www.caranddriver.com/news/a31098296/autonomous-cars-...

Even Tesla, already realized its not just a sensor problem, solving self-driving needs a higher level algo than can put the different sensors, in context of situational awareness. Note that in no any other way I would consider Tesla an example to follow ;-) And do not think they are any closer to getting a working system. Some of the statements show at least a second level understanding of what is required. Sensors are a means to it. Its about situational awareness but also inference.

"LIDAR is a fool’s errand… and anyone relying on LIDAR is doomed. — Elon Musk"

https://youtu.be/HM23sjhtk4Q

Re: Scenarios in which Tesla FSD Beta 9.0 fails

#113

Post by LeoPanthera » Mon, Jul 12, 2021, 6:44 PM UTC

Syonyk wrote: post #11

I don't understand how something this broken is allowed to operate on public roads. If I drove like Tesla's FSD seems to based on the videos I've seen, I'd be pulled out of my car and arrested on (well founded) suspicions of "driving while hammered." After a decade of work, it's not capable of doing much beyond "blundering around a city mostly without hitting stuff, but pay attention, because it'll try to hit the mos…

> I don't understand how something this broken is allowed to operate on public roads.

It's important to point out that this software is currently only offered as a private beta to deliberately selected testers. Now, maybe they shouldn't be using it on public roads either, but at least it's not available to the general public.

Re: Scenarios in which Tesla FSD Beta 9.0 fails

#114

Post by AndrewBissell » Mon, Jul 12, 2021, 6:47 PM UTC

Syonyk wrote: post #11

I don't understand how something this broken is allowed to operate on public roads. If I drove like Tesla's FSD seems to based on the videos I've seen, I'd be pulled out of my car and arrested on (well founded) suspicions of "driving while hammered." After a decade of work, it's not capable of doing much beyond "blundering around a city mostly without hitting stuff, but pay attention, because it'll try to hit the mos…

You would think that just the name "Tesla Full Self Driving Beta 9.0" would be giving people some pause here.

Re: Scenarios in which Tesla FSD Beta 9.0 fails

#115

Post by rasz » Mon, Jul 12, 2021, 6:47 PM UTC

serverholic wrote: post #72

Syonyk wrote: post #11

I don't understand how something this broken is allowed to operate on public roads. If I drove like Tesla's FSD seems to based on the videos I've seen, I'd be pulled out of my car and arrested on (well founded) suspicions of "driving while hammered." After a decade of work, it's not capable of doing much beyond "blundering around a city mostly without hitting stuff, but pay attention, because it'll try to hit the mos…

We are going to need a fundamental breakthrough in AI to achieve the level of FSD that people expect. Like a convolutional neural network level breakthrough.

Most problems in those clips didnt require general AI, they were caused by shit vision algorithms. Car didnt spot huge ass monorail columns ...

Re: Scenarios in which Tesla FSD Beta 9.0 fails

#116

Post by LeoPanthera » Mon, Jul 12, 2021, 6:48 PM UTC

zyang wrote: post #16

It appears Tesla FSD just ignores things it doesn't understand, which is really dangerous in a production vehicle.

The vehicle is production but this particular software is not, it's a private beta and not available to the general public.

Re: Scenarios in which Tesla FSD Beta 9.0 fails

#117

Post by ectopod » Mon, Jul 12, 2021, 6:50 PM UTC

tshaddox wrote: post #30

ethbr0 wrote: post #9

Earlier quoted context omitted.

Safely operating a motor vehicle via a steering wheel, accelerator, and brake (and sometimes clutch and gearshift) is a completely different skillset than monitoring an automated system in realtime. Novel skill requirements include: interpreting FSD UI, anticipating common errors, recognizing intent behind FSD actions, & remaining vigilant even during periods of autonomous success.

No, it’s just unsafe (and presumably illegal in most places) to operate a motor vehicle without maintaining constant awareness of the road and control of the vehicle. There’s no additional training or license that permits you to stop maintaining that awareness and control, just like there’s no additional training that permits you to drive under the influence of alcohol.

I agree, but if you have to maintain constant control of the vehicle then it isn't self-driving.

Re: Scenarios in which Tesla FSD Beta 9.0 fails

#118

Post by dragontamer » Mon, Jul 12, 2021, 6:51 PM UTC

belter wrote: post #87

serverholic wrote: post #72

Earlier quoted context omitted.

We are going to need a fundamental breakthrough in AI to achieve the level of FSD that people expect. Like a convolutional neural network level breakthrough.

I suggest some simple Smoke Tests. I love that concept for software testing. It could be applied here to test if we are getting closer or further from what humans can do. Some examples: Smoke Test 1: Driving with Snow Smoke Test 2: Driving with Rain Smoke Test 3: You are driving to a Red sign and you notice 50 meters ahead a pedestrian has its headphones on. The pedestrian is distracted and looking at traffic coming…

This "FSD" package is still at the "Don't hit this balloon we dressed up to look like a human" stage.

Re: Scenarios in which Tesla FSD Beta 9.0 fails

#119

Post by bob1029 » Mon, Jul 12, 2021, 6:53 PM UTC

serverholic wrote: post #72

Syonyk wrote: post #11

I don't understand how something this broken is allowed to operate on public roads. If I drove like Tesla's FSD seems to based on the videos I've seen, I'd be pulled out of my car and arrested on (well founded) suspicions of "driving while hammered." After a decade of work, it's not capable of doing much beyond "blundering around a city mostly without hitting stuff, but pay attention, because it'll try to hit the mos…

We are going to need a fundamental breakthrough in AI to achieve the level of FSD that people expect. Like a convolutional neural network level breakthrough.

I think we should first start with a fundamental breakthrough in our willingness as engineers to admit defeat in the face of underestimated complexity.

Once we take proper inventory of where we are at, we may find that we are still so far off the mark that we would be inclined to throw away all of our current designs and start over again from (new) first principles.

The notion that you can iteratively solve the full-self-driving problem on the current generation of cars is potentially one of the bigger scams in tech today (at least as marketed). I think a lot of people are deluding themselves about the nature of this local minima. It is going to cost us a lot of human capital over the long haul.

Being wrong and having to start over certainly sucks really badly, but it is still better than the direction we are currently headed in.

Re: Scenarios in which Tesla FSD Beta 9.0 fails

#120

Post by cs702 » Mon, Jul 12, 2021, 6:58 PM UTC

Many of the comments here seem a bit... unfair to me, considering that these clips were handpicked.

I watched (and fast-forwarded) through a few of the original, full-length videos from which these clips were taken. The full-length videos show long drives (in some cases, hours long) almost entirely without human intervention, under conditions explicitly meant to be difficult for self-driving vehicles.

One thing I really liked seeing in the full-length videos is that FSD 9 is much better than previous efforts at requesting human intervention in advance, with plenty of time to react, when the software is confused by upcoming road situations. The handpicked clips are exceptions.

For BETA software, FSD 9 is doing remarkably well, in my view. I mean, it's clearly NOT yet ready for wide release, but it's much closer than all previous versions of Tesla FSD I've seen before, and acceptable for a closed-to-the-public Beta program.