Discussion summary

Waymo reported teen riders for bad behavior, including using toy guns, and involved police. Discussions covered safety concerns, remote locking, and AI decision-making.

What the discussion says

  • Some users worry about safety and false positives leading to police involvement.
  • Concerns about remote locking and privacy with AI surveillance.
  • Fears about future AI judgment and potential misuse by hackers.
The car sent a report to police about toy projectile emitters.
sameers
Locking people in a car remotely raises safety and ethical questions.
kylecazar

Join the discussion

Write your take first — we'll ask for email only when you're ready to publish.

  • Hacker News
  • One time I was settling into a long Waymo ride (about 15 miles through traffic) and it was rolling down a wide urban street where a white English-speaking guy like me is distinctly a minority.

    And I was playing Pat Benatar's "Hit Me with Your Best Shot" with the windows wide open, and waving at other cars, and singing along at the top of my lungs, and eventually, Waymo Support rings in, and the guy goes, "just checking in! Is everything alright there?" and I go "er... yeah" and he goes "OK! thank you bye!"

    So I decided maybe there had been an external complaint, and I should perhaps keep the windows up, and the music down a few notches from now on. Especially as a minority in such a neighborhood like that one.

  • Ha! I don't think there was an external complaint. I suspect that was the mic/cameras IN the car alerting support because you were loud and they wanted to make sure you were okay. :)
  • https://archive.ph/yUFqG

    The car sent the San Mateo PD a report that the riders had toy projectile emitters (aka "guns," I suppose) they were using to shoot pellets at passers-by. It pulled into a parking lot until cops arrived. It doesn't report if it locked the riders in.

    I was just thinking I want to try at least one joyride but no, now I'm not even interested in giving them those $20.

  • hmmm kidknapped due to a false positive would be a bit of an oops to try and rewind. its bad enough that the cops do it [just try and get them to give you a ride back to where thay took you from, even when they are told to cut you loose]

    it causes real damages when you basically abduct and strand someone.

    these "kids" were shooting orbees out the windows, so mischief on them, but get it wrong once and it could get expensive.

  • Even if there was a way to remotely engage child lock, the front doors are still always unlocked from the inside.
  • I suspect they weren't locked in and stayed because they didn't know the police were coming to their location.

    Locking people in a car, remotely, as a private company, for "bad behavior", would be a mess.

  • My main concern is the deceit, "[s]topping for mechanical issues". Nah bruh
  • In my country (New Zealand) this would be very illegal

    You cannot detain someone because you do not like their behavior, in almost all situations. In civilised countries

    It is called kidnapping

  • As Benn Jordan pointed out, imagine what a hacker group or ICE could do with this. Fun times!
  • AI is just a tool. The real responsibility belongs to the developers, companies, and users who decide how it's built and used—not the AI itself. Anyway, that's insane. But if they were trying to damage the car, I can understand why Waymo reported them to the police.
  • I'm 99% sure that a human made these decisions after viewing the cameras in the vehicle, not that the Waymo vehicle itself decided to contact the police and park in a parking lot with these kids.

    This honestly seems pretty ideal. In a dangerous situation, the only risk is property damage - there's no driver to threaten!

  • Your view is dangerous because you don't see how this can be used to harm you.
    by 4d4m
  • > This honestly seems pretty ideal.

    I disagree. A decent response would have been for the car to simply refuse to continue the ride rather than kidnapping the ne'er-do-wells.

    That Waymo did this means that the hesitations I had about using the service are amplified by a few orders of magnitude. I'm not willing to risk someone at Waymo (or, worse, a machine at Waymo) making the wrong call and kidnapping me by mistake.

  • Agree. AI is not there yet, but it seems it is coming, and that would be scary. People will one day be in the scenario, where AI judges human behavior, and then decides what type of punishments to dish out based on it.