News for hackers. Stuff that matters.

A comment on Self-driving cars are headed toward an AI roadblock.

Re:Self-driving cars are headed toward an AI roadblock by adrianN on Thursday July 05, 2018 @12:24PM (#17464514)
The crashes that the article claims came from edge-cases that the engineers didn't account for weren't, in fact, edge-cases. The Tesla crashes are due to the way Tesla's system works, it doesn't recognize stationary hazards. Similarly Uber's system failed to brake for a hazard that it could clearly see. That's not an edge-case, it doesn't have to know that the object is a woman pushing a bicycle. It's a solid object moving steadily into the car's trajectory.
[ Parent | on HN ]
Re:Self-driving cars are headed toward an AI roadblock by gwern on Thursday July 05, 2018 @03:03PM (#17465863)
Citing the Uber case is particularly egregious. Have we all forgotten the NTSB report? It was only a few weeks ago. The Uber car tried to brake, but Uber had deliberately disabled the brakes for a more comfortable ride.
[ Parent | on HN ]
Re:Self-driving cars are headed toward an AI roadblock by UncleEntity on Thursday July 05, 2018 @05:19PM (#17466935)
Which basically makes the argument, to provide "a more comfortable ride" they had to disable a critical safety system that didn't do its job good enough to be used all the time. There were so many false positives that it made the car a road hazard due to random hard breaking.
[ Parent | on HN ]
Re:Self-driving cars are headed toward an AI roadblock by kwhitefoot on Friday July 06, 2018 @02:12AM (#17469322)
> There were so many false positives that it made the car a road hazard due to random hard breaking.

Do you know that for certain? Or is it just that Uber were too lazy to cope with any complications that might arise if Volvo's system braked?

[ Parent | on HN ]
Re:Self-driving cars are headed toward an AI roadblock by brookhaven_dude on Thursday July 05, 2018 @12:47PM (#17464748)
Edges cases are what really matters in this case!

Isn't there a saying that 80% of the code is written to handle 20% of the cases?

[ Parent | on HN ]
Re:Self-driving cars are headed toward an AI roadblock by ltdanimal on Thursday July 05, 2018 @02:37PM (#17465659)
I think you have that flipped. 20% of code covers 80%. As you try to get to 100% coverage of all possible scenarios the complexity gets in some cases exponentially harder as all the low hanging fruit is picked
[ Parent | on HN ]
Re:Self-driving cars are headed toward an AI roadblock by the-dude on Thursday July 05, 2018 @03:37PM (#17466151)
So he is exactly right : 80% of the code handles edges cases. You are the one who is flipped.
[ Parent | on HN ]
Re:Self-driving cars are headed toward an AI roadblock by cozzyd on Thursday July 05, 2018 @05:53PM (#17467188)
20% of code covers 80%. 80% of code covers 19%.
[ Parent | on HN ]
Re:Self-driving cars are headed toward an AI roadblock by Stanleyc23 on Thursday July 05, 2018 @01:21PM (#17465041)
but the problem is that you need to properly resolve base cases before starting on edge cases. The fact that base case scenarios are not properly solved hints at major problems in execution.
[ Parent | on HN ]
Re:Self-driving cars are headed toward an AI roadblock by mdorazio on Thursday July 05, 2018 @03:22PM (#17466027)
I like to think of it as an order of magnitude increase for each 9. Ex. If getting to 90% coverage takes 10 hours, getting to 99% will take 100, getting to 99.9% will take 1000, etc.
[ Parent | on HN ]
Re:Self-driving cars are headed toward an AI roadblock by the-dude on Thursday July 05, 2018 @02:49PM (#17465764)
Not specific to code : that is the Pareto principle.
[ Parent | on HN ]