Comment by almosthere

Comment by almosthere 3 days ago

20 replies

View on Hacker News

I don't think it makes sense to lump some drivers better than waymo and worse than waymo. A human brain automatically thinks of all the scenarios, where Waymo has pre-programmed ones (and some NN based ones). So it's scenarios by scenario.

Consider this scenario:

5 kids are walking on the sidewalk while you're driving past them. But suddenly a large dumpster is blocking your view of them just as you pass. You saw them before the dumpster, but not after your car and the dumpster completely blocks the view.

Does a human brain carry some worry that they suddenly decide to run and try to cross the street after the dumpster? Does Waymo carry that worry or just continue to drive at the exact same speed.

Again, it's not like every driver will think about this, but many drivers will (even the bad ones).

jobs_throwaway 3 days ago

> A human brain automatically thinks of all the scenarios

I don't think this is true. There are infinitely many scenarios in a complex situation like a road with traffic, cars parked, pedestrians about, weather, etc. My brain might be able to quickly assess a handful, but certainly not all.

Reply View 14 replies
  • xorbax 3 days ago

    > like a road with traffic, cars parked, pedestrians about, weather

    Not all of those need to be done "quickly". That's where LLMs fail

    You note the weather when you leave. You understand the traffic five minutes ahead. You recognize pedestrians far ahead of time.

    Computers can process a lot in fractions of a second. Humans can recognize context over many minutes.

    The Waymo may have done better in the fraction of a second, but humans can avoid being in that situation to begin with.

    Reply View 2 replies
    • jobs_throwaway 3 days ago

      Computers can take all of those things into account as well

      Reply View 1 reply
      • SirMaster 2 days ago

        Can, but don't.

        It doesn't seem like self driving cars take into account the icy conditions of roads for one simple example.

        Reply View 0 replies
  • jjav 3 days ago

    There aren't infinitely many scenarios to consider, but even if that's a figure of speech, there aren't thousands or even hundreds.

    If there's ten kids nearby, that's basically ten path scenarios, and that might be reduced if you have great visibility into some of them.

    > My brain might be able to quickly assess a handful, but certainly not all.

    What would you do if you can't assess all of them? Just keep driving same speed?

    If the situation is too overwhelming you'll almost certainly back off, I know I would. If I'm approaching that school block and there's like 50 small kids running around in all directions, I have no idea what's going on and who is going where, so I'm going to just stop entirely until I can make some sense of it.

    Reply View 7 replies
    • robotresearcher 3 days ago

      > here aren't infinitely many scenarios to consider, but even if that's a figure of speech, there aren't thousands or even hundreds.

      There are a very, very large number of scenarios. Every single possible different state the robot can perceive, and every possible near future they can be projected to.

      Ten kids is not 10 path scenarios. Every kid could do a vast number of different things, and each additional kid raises the number of joint states to another power.

      This is trivially true. The game that makes driving possible for humans and robots is that all these scenarios are not equally likely.

      But even with that insight, it’s not easy. Consider a simple case of three cars about to arrive at an all-way stop. Tiny differences in their acceleration - potentially smaller differences than the robot can measure - will result in a different ordering of cars taking turns through the intersection.

      It’s a really interesting problem.

      Reply View 2 replies
      • jjav 2 days ago

        > Ten kids is not 10 path scenarios. Every kid could do a vast number of different things, and each additional kid raises the number of joint states to another power.

        This is the difference between computing and humans. The car will attempt to compute all possible path scenarios because it has no instict, and it might not be possible to compute everything in real time so it might fail.

        But the human will easily deal with the situation.

        Try running through a sports field in an elementary school during lunch, full of unpredictable kids running around. Can you make it from one side to the other without crashing into a whole bunch of kids? Of course you can. You didn't need to try to compute an exponential number of scenarios, you just do it easily. The human brain is pretty amazing.

        Reply View 1 reply
        • robotresearcher 2 days ago

          In fact no computer approach attempts to compute all possible path scenarios since we know that’s not tractable.

          And current practical approaches are mostly end to end (or nearly) ML systems that do not compute a lot of alternative paths, and they work in approximately constant time independent of the scenario.

          You strongly imply that computers can’t drive, but you could have written that in a Waymo.

          Reply View 0 replies
    • lillecarl 3 days ago

      It should be trivial for Waymo to implement a "drive carefully near schools" feature, and if really spicy "drive REALLY carefully near schools at these times" feature.

      Safe driving starts with speed, lowering speed and informing the passengers seems like a no-brainer.

      Reply View 1 reply
      • jondwillis 2 days ago

        Feels like bitter lesson fodder to special case things like this

        Reply View 0 replies
    • jobs_throwaway 3 days ago

      It was a figure of speech, but I think you're undercounting. When you consider interactions between all the things, even with just a handful of variables (and I think there are many more than a handful) you get a huge number of scenarios.

      Reply View 1 reply
      • Mawr 2 days ago

        But you can group them into categories that can all be handled the same. For example, a child darting from behind a vehicle is a line-of-sight issue.

        To fix that, you program the car to handle situations with obstructed vision, which will handle not just this specific scenario, but all relating to obstructed line-of-sight — basically slow down enough to be able to stop in time in case something jumps out from behind the obstacle.

        Really though, this is less of an engineering problem and more of a social cost-benefit analysis one.

        On one hand, I'd say hitting a kid at 6mph in the worst case scenario once in a blue moon probably isn't that big of a deal.

        On another, someone here calculated that "even 1MPH slower would likely have resulted in no contact in this scenario".

        So really, it's not possible to say whether this was handled properly or not without access to data only Waymo has and establishing some standard of how much injury we're okay with vs the impact on travel times. Remember, we're seemingly ok with ~40 000 americans dying every year due to car transportation.

        Reply View 0 replies
  • robotresearcher 3 days ago

    This is the classical ‘Frame Problem” of AI. How do you consider, even if only to reject, infinite scenarios in finite time? Humans and other animals don’t seem to suffer from it.

    Reply View 0 replies
  • almosthere 3 days ago

    God I wish I re-read my statement, I was more focused on Humans think of an unlimited number of scenarios - not necessarily all. A computer will only think of pre-programmed ones.

    Reply View 1 reply
    • estearum 3 days ago

      The computer isn't pre-programmed though. These computers are trained similar to how human brains are (though obviously brains are still vastly, vastly, vastly superior to computers for tasks like this).

      Reply View 0 replies
IAmBroom 3 days ago

> A human brain automatically thinks of all the scenarios, ...

Patently, obviously false. A human brain will automatically think of SOME scenarios. For instance, if a collision seems imminent, and the driver is holding a cup of coffee, these ideas are likely to occur to the driver:

IF I GRAB THE STEERING WHEEL AND BRAKE HARD, I MIGHT NOT HIT THAT PEDESTRIAN IN FRONT OF ME.

IF I DON'T CONTINUE HOLDING THE COFFEE CAREFULLY, I MIGHT GET SCALDED.

THIS SONG ON MY RADIO IS REALLY ROCKING!

IF I YANK MY WHEEL TO THE LEFT, I MIGHT HIT A CAR INSTEAD OF A HUMAN.

IF I BRAKE HARD OR SWERVE AT ANY TIME IN TRAFFIC, I CAN CAUSE AN ACCIDENT.

Experiments with callosal patients (who have damaged the connective bridge between the halves of their brains) demonstrate that this is a realistic picture of how the brain makes decisions. It offers up a set of possible actions, and attempts to choose the optimal one and discard all others.

A computer program would do likewise, EXCEPT it won't care about the coffee cup nor the radio (remove two bad choices from consideration).

It still has one bad choice (do nothing), but the SNR is much improved.

I'm not being hyperbolic; self-preservation (focusing on keeping that coffee in my hand) is a vital factor in decision-making for a human.

> ...where Waymo has pre-programmed ones (and some NN based ones).

Yes. And as time goes on, more and better-refined scenarios will be added to its programming. Eventually, it's reasonable to believe the car software will constantly reassess how many humans are within HUMAN_RUN_DISTANCE + CAR_TRAVEL_DISTANCE in the next block, and begin tracking any that in an unsafe margin. No human on Earth does that, continually, without fail.

> Does a human brain carry some worry that they suddenly decide to run and try to cross the street after the dumpster? Does Waymo carry that worry or just continue to drive at the exact same speed.

You continue to imply that Waymo cannot ever improve on its current programming. Does it currently consider this situation? Probably not. Will it? Probably.

Reply View 1 reply
  • almosthere 3 days ago

    God I wish I re-read my statement, I was more focused on Humans think of an unlimited number of scenarios - not necessarily all. A computer will only think of pre-programmed ones.

    Reply View 0 replies
dmd 3 days ago

You are vastly overestimating most drivers. Most drivers aren't even looking out the window the majority of their time driving.

Reply View 0 replies
shadowgovt 3 days ago

For what it's worth, that kind of lumping of drivers is more-or-less one of the metrics Waymo is using to self-evaluate. Perfect safety when multi-ton vehicles share space with sub-300-pound humans is impossible. But they ultimately seek to do better than humans in all contexts.

Reply View 0 replies
assaddayinh a day ago

Didnt the lift car that ran that woman over had this exact problem because without the memento dementia it would have driven to slowly.

Reply View 0 replies