Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Tesla AI Day [video]

livestream.tesla.com

Post reply on HN
Print view View original

131–140 of 187 posts

Re: Tesla AI Day [video]

#131

Post by goldbattle »

2bitencryption wrote: post #28

A big part of the presentation so far was an engineer describing the huge difficulty of stitching together the multiple cameras into one vector space that can be the input to the network, instead of treating each camera individually. Seems like the biggest problem was each pixel from a camera does not tell you how far away it is, so even if you know the camera is X feet off the grounding pointing at Y degrees, you do…

While taking two sequential images from the same camera in can provide this depth information also. But simple use of stereo cameras can solve a magnitude of problems (standing still and low parallax motions). Traditional stereo and even machine-learning based methods have have great success and accuracy for many years and could easily be an alternative to LIDAR also.

I really don't know why this isn't leveraged more (maybe it is and I am unaware?).

Re: Tesla AI Day [video]

#132

Post by ScionOfDesign »

blahblahblogger wrote: post #129
bitsoda wrote: post #110

Earlier quoted context omitted.

The whole presentation is basically a recruiting event.

Honest Q for anyone here, how is Tesla's software-eng culture? I've of course heard of people on the ground floor of Tesla (factory workers) not being treated well, I've heard of SpaceX employees having high churn as they get "burnt through", and of Elon being kind of an ass. But having said that I have a Tesla and am one of those types that like to work on things that bring me value or that I can have some sort of t…

I did some contract work for Tesla last year over the course of ~8 months. If I were to pick a few key words it would be: High pressure, somewhat disorganized, and results-driven.

It was high pressure in that there were hard deadlines, and these deadlines needed to be met. If they weren't met, you needed to have a damn good explanation for as to why they weren't met.

Somewhat disorganized in that they did have a complete dev-ops lifecycle, but it felt kind of patched together. The team I was on did not do any formal standups or sprint planning. There was a very robust QA, validation, and homologation process though. They expected developers to do most of the QA, though. The point of the actual QA team was to ensure that there were 'no surprises'. I happened to be paired with a dedicated QA rep, though, which was nice.

Results-driven in that you were expected to do what you needed to do to get your job done. "Nobody told me what to do" was not an excuse. If you needed something done by someone else, you were expected to take initiative and ask and bug people until it got done. Most employees knew this, however, and were happy to help.

Engineers somewhat idolized Elon but also legitimately did not want to be the one to tell him bad news (such as they would prefer a different company rep do it). There are a lot of really smart people at Tesla, and you really do get to work on things that have never been done before (I got plenty of news articles written about what I worked on), but you really need to be dedicated to the company.

When I was there, there was a lot of crunch towards the end of my contract, and they said it wasn't usual, but I'd say expect crunch at least once or twice a year.

Re: Tesla AI Day [video]

#133

Post by dragontamer »

bane wrote: post #88

Bottom line, robots coming soon. He's slowly building out the roadmap for the labor pool, transportation, communications system and so on for the colonization of Mars.

They were supposed to do coast to coast self driving years ago according to these hype videos

https://www.digitaltrends.com/cars/tesla-coast-to-coast-auto...

Hard to take them seriously given how much hype they generate each year.

Re: Tesla AI Day [video]

#134

Post by sgustard »

tru3_power wrote: post #34
ModernMech wrote: post #30

Earlier quoted context omitted.

Yes. This is exactly what LIDAR is for. This is the reason every single team in the 2007 DARPA Urban challenge that finished the race was equipped with a Velodyne laser. We knew it was a critical enabling technology 14 years ago, and it baffles me Tesla is eschewing it. Tesla's lack of this technology is the main reason I feel they will never achieve what they claim to without a huge breakthrough in AI.

What was their reasoning for not using Lidar?

See, for example, Andrej's talk starting at the 6-minute mark. Lidar requires a pre-rendered detailed map of the lanes, traffic lights and obstacles. Vision can operate in any novel environment the car is not preprogrammed for and is thus more scalable.

https://www.youtube.com/watch?v=a510m7s_SVI

Re: Tesla AI Day [video]

#135

Post by ajross »

AlotOfReading wrote: post #120
ajross wrote: post #26

Earlier quoted context omitted.

> At the end of the day though, there is no way for me or anyone to audit the neural networks or their training data. How can anyone trust them? By external measurements and statistics. You measure an ML engine by performance. Literally no one anywhere knows how to debug these things by looking inside the model. If it's a "fragile system" then it will be shown to be so in data. At the end of the data, either cars cra…

How can you measure the performance of an ML engine in untested conditions? Tesla's whole goal is to achieve L5 capabilities. That means anywhere in any conditions, which is clearly impossible to comprehensively validate with a finite test set. It's plausible that we could get a long period of acceptable safety before a black swan event suddenly causes deaths because it can't handle smoke, heavy rain/fog, or blizzard…

That's just ludditism. That logic works for any new technology. How do you know vaccines are safe under all conditions? How do you know planes won't crash under some "black swan" event? How do you know GM food won't poison you? How do you know your fuel tank won't explode? How do you know your ocean liner won't hit an iceberg and sink? There's literally no technology in the world that meets your standards.

You don't know. You measure and you decide, based on numbers and science and moral reasoning about risk. You don't get absolutes from anything else, why are you demanding it from cars?

Re: Tesla AI Day [video]

#136

Post by screye »

KETpXDDzR wrote: post #20
Invictus0 wrote: post #15

I like Andrej and I think he's a brilliant guy. At the end of the day though, there is no way for me or anyone to audit the neural networks or their training data. How can anyone trust them? I think this is really a fundamental problem with using neural networks in self-driving cars--I don't want to put my life in the hands of a fragile system that I, and no one alive, can understand. We are being asked to close our…

Compare it to a horse. No one knows what's going on in the horse's brain. However, we learned to trust it and how to deal with it safely. E.g., blinkers (the devices to limit a horse's peripheral view).

Horse collisions didn't usually end up in death. Horses also weren't sprinting at 65 mph through 8 lane highways.

Re: Tesla AI Day [video]

#138

Post by sangnoir »

panick21 wrote: post #123
c-cube wrote: post #108

Earlier quoted context omitted.

Just because not needing lidar would be great for Tesla, doesn't make it real. That's wishful thinking. Tesla is not entitled to a well working lidar-free solution just because they need to sell cars right now.

Can you point me to literally any evidence what so ever that the results fake? Is the FSD Beta regularly crashing into things it put at the wrong range? Tesla has cars with Lidar on the road to check the models as well. And who says anything about entitled, they have spent a huge amount of effort and therefore money into a solution that they think will be better. Nobody with Lidar has delivered either so what is bett…

> Is the FSD Beta regularly crashing into things it put at the wrong range?

As a matter of fact, yes: self-driving Teslas have been crashing into emergency vehicles so often, they are being investigated[1]

1. https://news.ycombinator.com/item?id=28197355

Re: Tesla AI Day [video]

#139

Post by FeepingCreature »

karpierz wrote: post #96
FeepingCreature wrote: post #56

Earlier quoted context omitted.

That isn't overfit, that's fit. Nothing can protect you if your training set just doesn't have any indication of the thing you want it to learn.

I didn't specify that the training set wasn't representative. All this shows is that you don't need parameters anywhere close to the number of test examples to overfit.

And my point is that is not what overfit is. Overfit is a specific problem where the network fails to recognize a commonality in the training set and instead interprets the irrelevant details of some subset of training samples (in the extreme, individual samples) as distinct properties.

Your example training set is not filled with noise that the network is picking up on to its detriment. Your example training set is simply not representative of the function you are trying to teach.

Re: Tesla AI Day [video]

#140

Post by tablespoon »

abc_lisper wrote: post #74
ModernMech wrote: post #30

Earlier quoted context omitted.

Yes. This is exactly what LIDAR is for. This is the reason every single team in the 2007 DARPA Urban challenge that finished the race was equipped with a Velodyne laser. We knew it was a critical enabling technology 14 years ago, and it baffles me Tesla is eschewing it. Tesla's lack of this technology is the main reason I feel they will never achieve what they claim to without a huge breakthrough in AI.

Idk.. animals seem to do fine without a lidar

> Idk.. animals seem to do fine without a lidar

Animals also have (non)-artificial general intelligence. That doesn't mean Tesla's making a reasonable design choice if they build something in a way that relies on having that kind of capability.

Post reply on HN