Comment on In Leaked Audio, Amazon Cloud CEO Says AI Will Soon Make Human Programmers a Thing of the Past

• Num10ck@lemmy.world ⁨3⁩ ⁨weeks⁩ ago

  there will be a massive building in like india with many thousand of atrociously paid workers donning VR goggles who spend their long hours constantly Quantum Leap finding themselves in traumatizing last second emergency situations that the AI gives up on. Instantly they slam on the brakes as hard as they can. They drink tea. there’s suicide netting everywhere. they were the lowest bidder this quarter.

  source

  • arken@lemmy.world ⁨3⁩ ⁨weeks⁩ ago

    I wish I could give this comment more than a simple upvote. I want to mail you a freshly baked cinnamon bun.

    source
• cogitase@lemmy.dbzer0.com ⁨3⁩ ⁨weeks⁩ ago

  Their accident rate continues to decrease and things like quorum sensing and platooning are going to push them to be better than humans. You’re never going to have a perfect system that never has accidents, but if you’re substantially better than humans in accidents per mile driven and you’re dramatically improving throughput and reducing traffic through V2X, it’s going to make sense to fully transition.

  I imagine some east Asian countries will be the first to transition and then the rest of the world will begrudgingly accept it once the advantages become clear and the traditional car driving zealots die off.

  source

  • AtomicTacoSauce@lemmy.world ⁨3⁩ ⁨weeks⁩ ago

    The robot taxi from Total Recall came to mind while reading your reply. Our future is almost assuredly dystopian.

    source
• KevonLooney@lemm.ee ⁨3⁩ ⁨weeks⁩ ago

  Plus, as soon as the cars can drive themselves people will stop needing Uber in many cases.

  No parking? Just tell your car to go park on a street 10 blocks away.

  Drunk? Car drives itself while you sleep.

  Going to the airport? Car drops you off and returns home. Car also picks you up when you are back.

  This is combined with the fact that people will do more disgusting things in an Uber without the driver there. If you have ever driven for Uber, you know that 10% of people are trying to eat or drink in the car. They are going to spill and it’s going to end up like the back of a bus.

  source

  • yildolw@lemmy.world ⁨3⁩ ⁨weeks⁩ ago

    Not sure if we’re agreeing and saying exactly the same thing here, but Uber’s business model is to get suckers who are bad at math to own the cars. Uber’s business model does not work if they have to own their own cars. Self-driving Uber doesn’t work because Uber would have to own the cars.

    source
• Nollij@sopuli.xyz ⁨3⁩ ⁨weeks⁩ ago

  “handle” is doing a lot of heavy lifting there. The signs are already there that all of these edge cases will just be programmed as “safely pull over and stop until conditions change or a human takes control”. Which isn’t a small task in itself, but it’s a lot easier than figuring out to continue (e.g.) on ice.

  source
• pbbananaman@lemmy.world ⁨3⁩ ⁨weeks⁩ ago

  Just like all humans can do right now, right?

  I never see any humans on the rode staring at their phone and driving like shit.

  source

  • Wilzax@lemmy.world ⁨3⁩ ⁨weeks⁩ ago

    The problem with self-driving cars isn’t that it’s worse than human drivers on average, it’s that it’s SO INCREDIBLY BAD when it’s wrong that no company would ever assume the liability for the worst of its mistakes.

    source

    • pbbananaman@lemmy.world ⁨3⁩ ⁨weeks⁩ ago

      But if the average is better, then we’re will clearly win by using it. I’m not following the logic of tracking the worst case scenarios as opposed to the average.

      source

      • Wilzax@lemmy.world ⁨3⁩ ⁨weeks⁩ ago

        Average is better means fewer incidents overall. But when there are incidents, the damages for those incidents tend to be much worse. This means the victims are more likely to lawyer up and go after the company responsible for the AI that was driving, and that means that the company who makes the self-driving software better be prepared to pay for those worst case scenarios, which will now be 100% their fault.

        Uber can avoid liability for crashes caused by their human drivers. They won’t be able to do the same when their fleet is AI. And when that happens, AI sensibilities will be measured my human metrics because courts are run by humans. The mistakes that they make will be VERY expensive ones, because a minor glitch can turn an autonomous vehicle from the safest driving experience possible to a rogue machine with zero sense of self-preservation. That liability is not worth the cost savings of getting rid of human drivers yet, and it won’t be for a very long time.

        source
• rottingleaf@lemmy.world ⁨3⁩ ⁨weeks⁩ ago

  Those self-driving cars are called trains. They already can be self-driving. In a situation where the computational complexity and required precision are somewhat controlled, that is, on train tracks.

  source