5 ms·

I actually think driverless cars will do better here. They will likely be able to recognize cars deviating from their current path in O(hundreds of milliseconds

by foota 7y ago

I actually think driverless cars will do better here. They will likely be able to recognize cars deviating from their current path in O(hundreds of milliseconds) and avoid a collision, whereas if a car from the opposite direction lane starts drifting into mine seconds before we hit there's little I can do.

beerandt 7y ago
I mentioned this elsewhere, but I'll elaborate- I don't think the detection or reaction time is the problem, it's the concept of useful recognition (and prediction). Is that first movement/deviation (that an AI car can certainly measure sooner than I can see) going to be a swerve within their lane or is the car going to come all the way over? It's not the watching for the trigger, but the intuition of what the trigger is, well, triggering. Humans have an uncanny ability in predicting the overall movement before any perceivable deviation. Of course I'm not saying it's magic; there's obviously some kind of signal to our subconscious that we're able to process and make useful. Perhaps AI can pick up on some of that. But it's not necessarily the "first movement" of the swerve that is most useful. And the "signal" that humans pick up on may not even be something externally measurable by the car. Like knowing of a particular interstate exit that drivers always seem to almost miss, but swerve across lanes to catch at the last second. There are also signals that are externally measurable, but have to be learned (seen on a previous trip, or warned by a friend). Like knowing about a bad pothole ahead, and knowing that the car in front of you is going to turn the corner, slam on the brakes and then swerve to avoid it. These are the kinds of things that I suspect AI will struggle to keep up with in the foreseeable future. And humans rely on this type of intuition way more than I think most people realize.
- jodrellblank 7y ago
  Is this the 'disconnected lever' problem of AI, applied to vehicle driving? Of course I'm not saying it's magic; there's obviously some kind of signal to our subconscious that we're able to process and make useful. That signal is connected to some much larger dataset of years of human experience - things like what a distracted driver might do, what a tired driver might do, what an old unmaintained vehicle looks like, what hints at an out-of-state or out-of-country driver who might not be aware of local routes or problems, how people tend to behave differently at dawn/rush hour/day/night. An AI car should be able to tell which is a busy junction pretty easily, but watching a car swerve in a way that might indicate the driver is not paying full attention will make you more cautious around that car for a long time up ahead. Still, AI cars which don't tailgate, will get an ENORMOUS safety advantage over the people who drive 2ft behind me at highway speeds. That alone might be enough to overcome every other benefit. "AI car doesn't get involved in multi-car pileup", "AI car didn't get shunted at traffic lights", "AI car let me in at that junction" will keep happening and keep boosting people's confidence in them little by little.
  - beerandt 7y ago
    >disconnected lever Yes, it's good to know this has a name. Thanks. This idea, but also extended to include the fact that as AI drives more, humans drive less, and might lose some of that knowledge. If AI is driving 80% if the time, but can't handle 20% of situations, then humans are doing the 20% most difficult driving with 80% less practice. Obviously this is over simplifying things, as lack of fatigue and other things will come into play. But I'm merely stating that both sides of the equation will likely change, not just one. >Still, AI cars which don't tailgate...ENORMOUS safety advantage As long as they do this in a way that doesn't create congestion, I agree. But I'm not yet convinced. Overly conservative driving can be just as, if not more dangerous than some aggressive driving. A lot of the risk might be merely transferred, not reduced. Preventable congestion is the bigger problem, tailgating is usually just a symptom. (Not saying it's ok, just that it's preventable in other, more indirect ways.) And many solutions to congestion require complicated risk-now vs risk-later evaluations (including some that you listed- id'ing a bad driver or a tire about to blow out). These require more decisive, if not aggressive, driver actions, like speeding for 20 seconds to get away from the bad driver or to clear the passing lane. Or accelerating aggressively on an on-ramp (and not have to stop at the end of it to wait for an opening, which is part of why on-ramp metering has become so popular). Or if you're on the interstate and it's posted for 45mph, but everyone else is driving 80mph. It's probably safer for you to speed (at least 60+) than to drive 45mph. That's not really compatible with AI's current approach of being overly conservative with everything. All this to say, I don't doubt AI will make some aspects of driving safer, but I still have doubts about whether it will improve overall safety, at least in the foreseeable future. Driving and traffic systems are complex.
LoSboccacc 7y ago
unless they're white trucks, it seems

by foota 7y ago

beerandt 7y ago

I mentioned this elsewhere, but I'll elaborate- I don't think the detection or reaction time is the problem, it's the concept of useful recognition (and prediction). Is that first movement/deviation (that an AI car can certainly measure sooner than I can see) going to be a swerve within their lane or is the car going to come all the way over? It's not the watching for the trigger, but the intuition of what the trigger is, well, triggering. Humans have an uncanny ability in predicting the overall movement before any perceivable deviation. Of course I'm not saying it's magic; there's obviously some kind of signal to our subconscious that we're able to process and make useful. Perhaps AI can pick up on some of that. But it's not necessarily the "first movement" of the swerve that is most useful. And the "signal" that humans pick up on may not even be something externally measurable by the car. Like knowing of a particular interstate exit that drivers always seem to almost miss, but swerve across lanes to catch at the last second. There are also signals that are externally measurable, but have to be learned (seen on a previous trip, or warned by a friend). Like knowing about a bad pothole ahead, and knowing that the car in front of you is going to turn the corner, slam on the brakes and then swerve to avoid it. These are the kinds of things that I suspect AI will struggle to keep up with in the foreseeable future. And humans rely on this type of intuition way more than I think most people realize.

jodrellblank 7y ago
Is this the 'disconnected lever' problem of AI, applied to vehicle driving? Of course I'm not saying it's magic; there's obviously some kind of signal to our subconscious that we're able to process and make useful. That signal is connected to some much larger dataset of years of human experience - things like what a distracted driver might do, what a tired driver might do, what an old unmaintained vehicle looks like, what hints at an out-of-state or out-of-country driver who might not be aware of local routes or problems, how people tend to behave differently at dawn/rush hour/day/night. An AI car should be able to tell which is a busy junction pretty easily, but watching a car swerve in a way that might indicate the driver is not paying full attention will make you more cautious around that car for a long time up ahead. Still, AI cars which don't tailgate, will get an ENORMOUS safety advantage over the people who drive 2ft behind me at highway speeds. That alone might be enough to overcome every other benefit. "AI car doesn't get involved in multi-car pileup", "AI car didn't get shunted at traffic lights", "AI car let me in at that junction" will keep happening and keep boosting people's confidence in them little by little.
- beerandt 7y ago
  >disconnected lever Yes, it's good to know this has a name. Thanks. This idea, but also extended to include the fact that as AI drives more, humans drive less, and might lose some of that knowledge. If AI is driving 80% if the time, but can't handle 20% of situations, then humans are doing the 20% most difficult driving with 80% less practice. Obviously this is over simplifying things, as lack of fatigue and other things will come into play. But I'm merely stating that both sides of the equation will likely change, not just one. >Still, AI cars which don't tailgate...ENORMOUS safety advantage As long as they do this in a way that doesn't create congestion, I agree. But I'm not yet convinced. Overly conservative driving can be just as, if not more dangerous than some aggressive driving. A lot of the risk might be merely transferred, not reduced. Preventable congestion is the bigger problem, tailgating is usually just a symptom. (Not saying it's ok, just that it's preventable in other, more indirect ways.) And many solutions to congestion require complicated risk-now vs risk-later evaluations (including some that you listed- id'ing a bad driver or a tire about to blow out). These require more decisive, if not aggressive, driver actions, like speeding for 20 seconds to get away from the bad driver or to clear the passing lane. Or accelerating aggressively on an on-ramp (and not have to stop at the end of it to wait for an opening, which is part of why on-ramp metering has become so popular). Or if you're on the interstate and it's posted for 45mph, but everyone else is driving 80mph. It's probably safer for you to speed (at least 60+) than to drive 45mph. That's not really compatible with AI's current approach of being overly conservative with everything. All this to say, I don't doubt AI will make some aspects of driving safer, but I still have doubts about whether it will improve overall safety, at least in the foreseeable future. Driving and traffic systems are complex.

LoSboccacc 7y ago

unless they're white trucks, it seems