Andrej Karpathy talks about how Tesla's NNs are structured and trained [video] • Hacker News Board

Re: Andrej Karpathy talks about how Tesla's NNs are structured and trained [video]

#191

Post by dzdt » Sun, Nov 10, 2019, 5:43 PM UTC

SloopJon wrote: post #166

Udik wrote: post #101

Earlier quoted context omitted.

What does Netflix do with NNs?

Their own page says, "our recommendation algorithms ... learning characteristics that make content successful ... optimize the production of original movies and TV shows ... optimize video and audio encoding, adaptive bitrate selection, and our in-house Content Delivery Network ... and advertising". https://research.netflix.com/research-area/machine-learning Here's another article on the subject: https://becominghuma…

Netflix put a lot of work into recommendation back in the dvd delivery days when their catalog was absolutely massive. Now in the streaming space the catalog they license is much smaller so recommendation is less important; basically they just advertise the popular stuff for your demographic.

Its a bit odd that legally one can rent out physical disks, but there is no corresponding way to legally get permission to rent out streaming content without negotiating with the rightsholder. But thats how it is...

Re: Andrej Karpathy talks about how Tesla's NNs are structured and trained [video]

#192

Post by adamnemecek » Sun, Nov 10, 2019, 6:00 PM UTC

Geee wrote: post #167

adamnemecek wrote: post #7

The trick for level 5 is learning the mapping between the lidar point cloud and the video stream. It’s the best of both worlds.

2D to 3D transform is simple trigonometry (using stereo / motion) and should be possible to learn without lidar. I think this is already a solved problem. One option though is to add lidars in random Teslas (e.g. 1/1000) to help with the labeling / learning.

The field is called photogrammetry. It's not...precise. Point clouds have a much richer structure.

Re: Andrej Karpathy talks about how Tesla's NNs are structured and trained [video]

#193

Post by kortex » Sun, Nov 10, 2019, 6:05 PM UTC

logicallee wrote: post #147

jacquesm wrote: post #105

The competition in this space is great but I can't help but wonder what would happen if instead all these companies pooled their resources and went after the goal collectively. There is so much duplication going on and the paths do not seem to me - as an outsider - to be all that divergent, which is usually a pre-condition for having a lot of independent efforts one of which will succeed. It's as if everybody wants t…

>Imagine every car manufacturer having a completely different take on what a car should be like from a safety perspective. We have standards bodies for a reason Roads are also governed by public bodies. Road signs are standardized and public. I think the government should take a much larger role in defining self driving cars. For example, rather than using computer vision to recognize signs, signs could be active sta…

> Would it be absurd for the crosswalk to know roughly how many people are at it, then broadcast this to the car

Yes, completely absurd. For one, many places are too sparse, poor, or unstandardized for this to be remotely economical. Two, people often don't cross at crosswalks (jaywalking). Level 5 AV is a thing where 99% coverage isn't good enough. You need a lot of nines.

That's why many think lidar systems are a crutch. If lidar can't work in snow or heavy rain (at least 1-2% of days in the north), then you need fallback which must still be >99% as effective to avoid incidents. But then why not just use the fallback?

Generating data? Sure. But for actual use in the data pipeline? Makes you rely on an ultimately untenable solution.

Re: Andrej Karpathy talks about how Tesla's NNs are structured and trained [video]

#194

Post by anticensor » Sun, Nov 10, 2019, 6:29 PM UTC

oska wrote: post #127

0-_-0 wrote: post #121

Earlier quoted context omitted.

Just replace "eight" with "two" and this could have been written about the human brain

And driving with only one is still legal.

Monocular or deaf driving is subject to many restrictions including being unable to drive a commercial vehicle.

Re: Andrej Karpathy talks about how Tesla's NNs are structured and trained [video]

#195

Post by eyeundersand » Sun, Nov 10, 2019, 6:40 PM UTC

protomikron wrote: post #130

Fun fact for all of you: Some time ago (around ~10 years) this guy (the presenter) was internet famous for being a Rubik cube speed solver and making tutorials and videos about that: https://www.youtube.com/watch?v=609nhVzg-5Q

Wow! Thank you.

What a blast from the past! I learned how to solve the Rubik's Cube blindfolded by watching him, back in the day. His tutorials are perfect and I've probably recommended his channel to ~50 people myself. Crazy he only has 36.6k subs.

Glad to see he's doing well.

Re: Andrej Karpathy talks about how Tesla's NNs are structured and trained [video]

#196

Post by p1esk » Sun, Nov 10, 2019, 6:53 PM UTC

jfoster wrote: post #173

jacquesm wrote: post #105

The competition in this space is great but I can't help but wonder what would happen if instead all these companies pooled their resources and went after the goal collectively. There is so much duplication going on and the paths do not seem to me - as an outsider - to be all that divergent, which is usually a pre-condition for having a lot of independent efforts one of which will succeed. It's as if everybody wants t…

> The competition in this space is great but I can't help but wonder what would happen if instead all these companies pooled their resources and went after the goal collectively. It would probably slow down. 9 women can't have a baby in 1 month. Besides that, the disagreements about approach, politics, or eventual competitive interests would probably bring things to a halt for a long time. I don't think the solutions…

Exactly. Look at Human Brain Project: $1B and 10 years later what exactly have they achieved? Just like you said - disagreements about approach, politics, or eventual competitive interests did bring things to a halt for a long time

Re: Andrej Karpathy talks about how Tesla's NNs are structured and trained [video]

#197

Post by leesec » Sun, Nov 10, 2019, 6:53 PM UTC

anonu wrote: post #177

londons_explore wrote: post #176

I'm still amazed that Teslas team isn't using a map... I know maps get outdated and are sometimes wrong, but having inaccurate knowledge of what's around the corner is far far more helpful than not having any clue whats around the corner. The smart solution would be to consider a map a probabilistic thing, which neural networks are really good at handling.

I'm still amazed Tesla has decided not to use lidar and instead just stick with cheap cameras. Better sensors are there, they're available, they're cheap and they can probably "see" better than plain old cameras... it doesn't make too much sense not to use them IMHO. But then again, I am not coding NNs for Tesla...

You're underestimating the cost issue. Obviously if it was literally 0% extra cost they would have a value benefit. The problem is making a great and cheap and profitable electric vehicle.

Re: Andrej Karpathy talks about how Tesla's NNs are structured and trained [video]

#198

Post by nielsole » Sun, Nov 10, 2019, 6:54 PM UTC

londons_explore wrote: post #187

anonu wrote: post #177

Earlier quoted context omitted.

I'm still amazed Tesla has decided not to use lidar and instead just stick with cheap cameras. Better sensors are there, they're available, they're cheap and they can probably "see" better than plain old cameras... it doesn't make too much sense not to use them IMHO. But then again, I am not coding NNs for Tesla...

While LIDAR's are certainly 'better' from a technological standpoint than not having anything, from a business standpoint it's less clear. LIDAR's are cheap, but not cheap enough yet to not seriously affect the bottom line if you put them into every car. It also will kill the resale price of cars without it, which in turn hurts the companies image and stock price.

If you are first to market with a level 4 autonomous taxi, unit economics will likely be great regardless of whether you put LIDAR or cameras in.

Re: Andrej Karpathy talks about how Tesla's NNs are structured and trained [video]

#199

Post by oneshot908 » Sun, Nov 10, 2019, 7:01 PM UTC

jfoster wrote: post #173

jacquesm wrote: post #105

The competition in this space is great but I can't help but wonder what would happen if instead all these companies pooled their resources and went after the goal collectively. There is so much duplication going on and the paths do not seem to me - as an outsider - to be all that divergent, which is usually a pre-condition for having a lot of independent efforts one of which will succeed. It's as if everybody wants t…

> The competition in this space is great but I can't help but wonder what would happen if instead all these companies pooled their resources and went after the goal collectively. It would probably slow down. 9 women can't have a baby in 1 month. Besides that, the disagreements about approach, politics, or eventual competitive interests would probably bring things to a halt for a long time. I don't think the solutions…

Also agreed, when hedge funds don't silo their quants, instead of seeing 50 different strategies from 50 quants, they get 50 variants of the same strategy, source:

https://www.amazon.com/gp/product/1119482089/ref=ppx_yo_dt_b...

Re: Andrej Karpathy talks about how Tesla's NNs are structured and trained [video]

#200

Post by mattrp » Sun, Nov 10, 2019, 7:02 PM UTC

iguy wrote: post #150

jacquesm wrote: post #113

Earlier quoted context omitted.

I'm reminded of the Manhattan project and I don't think they would have succeeded in their goal if they had tried to run 10 of those at once. There just aren't that many really great scientists in a space this narrow.

They built complete factories for three different technologies, and finally delivered two totally different bomb designs. Which is a sign they were pretty worried about choosing to put all their smart guys on one path, and picking the wrong one. And the number of smart guys they had was certainly a real constraint. Even under their very stressed circumstances, exploration and competition were important.

This is correct - read the fantastic “making of the atomic bomb” - the planners kept competition alive for as long as possible because no one really knew which path was going to work. In contrast the nazi project, among other reasons, failed because it focused nearly exclusively on heavy water and didn’t have internally competing projects.