3 ms·
I think that's a dangerous way to evaluate self-driving systems. With a little practice, any 10-year-old can "drive quite well" under limited conditions. But wh
by wpietri 6y ago
I think that's a dangerous way to evaluate self-driving systems. With a little practice, any 10-year-old can "drive quite well" under limited conditions. But what makes for a good driver is not the 99% of the time dealing with optimum conditions, but the how they handle the long tail of increasingly challenging/dangerous situations.
I suspect a lot of companies could deliver a Tesla-equivalent system. They're just opposed to killing people in the process. See, for example, this thread from a host of The Autonocast contrasting Tesla's approach with others: https://twitter.com/Tweetermeyer/status/1383778011766398983 https://twitter.com/Tweetermeyer/status/1383778011766398983
- fastball 6y agoLevel 4 doesn't actually include the long-tail though, that's Level 5. It's really a terrible benchmark system. Tesla could be level 4 tomorrow if they said (like Waymo) "we're gonna enable FSD only in the greater Phoenix area", where the roads are well-maintained, they could map them all out, etc. The issue here is that Tesla actually wants a generalized system that will work anywhere a human driver can drive, and does not seem interested in geofencing at all until they've reached that (level 5) point.
- debrice 6y agoI find this strategy very close to the lean startup model (MVP). Go to market with something you believe is better than what’s out there early, have rapid feedback and iterate. Feels like a risky bet for Waymo and others to reach level 4 or 5 only in a limited environment, you might end up with so many untested assumptions
- fastball 6y agoYes, to me this is also another reason Tesla decided to not go with LIDAR (besides cost): if your self-driving car uses a different sensing system than a human, there are no guarantees that this car will be able to drive anywhere a human can drive. Meanwhile, if your system is relying on entirely the same thing humans rely on, the likelihood seems higher you'll end up with a more generalized solution, since all roads / etc on earth are built around vision, not LIDAR.
- kmonsen 6y agoDo you have any evidence Tesla could do this in Phoenix? I think that is highly unlikely and at this stage might not be possible without lidar.
- fastball 6y agoWith a geofenced area you can hard-code behaviors in any area where FSD struggles. I see no reason to believe they couldn't do this – FSD can already drive quite a few places with zero assistance, and most of the hiccups seem to be around edge-cases, which are easily pre-programmed.
- wpietri 6y agoDo you have any evidence that's what Waymo is doing, hardcoding behaviors in specific spots/
- fastball 6y agoWhen I say "hardcoding" I don't actually mean hardcoding. I mean forcing the ML system to keep iterating on specific problems that it did poorly at, with human supervision. e.g. the car did the wrong thing at the intersection of 9th and Elm? Give it 100 similar situations (at 9th and Elm), let it generate lots of possible behaviors, and then have humans tell it which are correct. Then whenever a car is at 9th and Elm, it uses that training as a priority. For a relatively small geofenced area that is a reasonable approach and should work well. Unfortunately it doesn't scale to the whole country/world.
- wpietri 6y agoWhy doesn't it scale? That's basically how humans do it. We learn general-case rules and then we learn how to handle specific situations specifically. You can observe that just by driving through an unfamiliar city with a local. They'll feed you a stream of tricks for optimizing the drive.
- judge2020 6y agoMost humans do not, indeed, work well in those 1% conditions. If they wanted to avoid damage to their cars, other cars, or human death/injury, they would pull over or choose not to drive when the weather calls for bad driving conditions, or when drunk. I imagine Tesla and the rest will be on-par with humans in these driving conditions within 5 years, but it’s definitely fair to ask that they be 10x as safe as a regular driver which might take 15 or 25 years.
- debrice 6y agoI don’t believe lots of companies can deliver what Tesla offers, or if so, have the ethics of not delivering for the sake of saving lives. If anything, history showed that most car companies ethics is very questionable when competition is fierce. Unless I misinterpreted the recent Tesla safety report, the Tesla FSD seems statistically safer than human drivers (please correct me if I’m wrong).
- wpietri 6y ago> history showed that most car companies ethics is very questionable when competition is fierce Which history? Safety of cars in the US has been improving for nearly a century: https://commons.wikimedia.org/wiki/File:USA_annual_VMT_vs_deaths_per_VMT.png https://commons.wikimedia.org/wiki/File:USA_annual_VMT_vs_de... If that graph indicates car companies are willing to sacrifice lives to get an edge, I wonder what you think the opposite would look like.
- cycomanic 6y agoThe statistic is highly misleading. Yes if I only turn on a feature in the most safe situations (e.g. on the highway) it's going to get better safety statistics than general driving which includes non-safe situations.
- jjcon 6y agoI’m not sure a random portlander with no expertise on this tech who makes their money selling self published ebooks on why you should be afraid of Tesla is exactly the right person to include in this conversation.