Re:Uber Wasted $2.5B on Self-Driving Cars
by d1zzy on Monday September 28, 2020 @08:53PM
(#24623044)
I see people keep stating this but I'm not so sure about it. Do we know what critically important data Tesla cars collect at scale _and_ are able to send back to mothership and who pays for that bandwidth/traffic? To my mind most of the useful information (video recordings or pictures of moments when the AI driver wasn't sure about things and/or human intervened) seems like it would take a lot of bandwidth to send back from moving cars.
[ Parent
| on HN ]
Re:Uber Wasted $2.5B on Self-Driving Cars
by cycrutchfield on Monday September 28, 2020 @09:12PM
(#24623181)
Yes, according to Karpathy. The cars upload data (typically still images) which are then labeled by humans and used as training data to improve the models.
Re:Uber Wasted $2.5B on Self-Driving Cars
by ethbr0 on Monday September 28, 2020 @11:29PM
(#24623934)
The impression I got from GTC a few years ago was that they're probably doing synthetic graphics simulations, based on edge cases identified by real world data.
You don't really need 1000x real examples of a scenario, if one still image provides enough information for you to mimic it in simulation. (E.g. toppled barrel, weathered paint color and reflective stripe, in front of exit guard rail)
Re:Uber Wasted $2.5B on Self-Driving Cars
by cycrutchfield on Tuesday September 29, 2020 @03:34AM
(#24625057)
I don’t have the link, but watch Karpathy’s recent talk on this topic on YouTube. He goes into detail about how this works. I’m not sure what their approach was years ago but it has changed significantly since.