Hacker News new | ask | show | jobs
by tmcw 1434 days ago
Sure, so the most direct answer is that a belief in self-driving cars that are safe and reliable and accessible isn’t grounded in any current evidence. We’ve spent a few billion on development and spun up dozens of companies and so far accomplished - relative to expectations - almost nothing. No self-driving trip across the country, like Musk thought we’d do years ago. The self-driving fleet in SF keeps breaking down.

Maybe in the future it’ll suddenly go from bad technology to good, but in my mind it’s like betting on generalized AI or small-scale cold-fusion nuclear: it could happen, but not a great idea to bet the farm on it.

Second. Car form factor is inefficient. Are we doing 1-seaters? 4 seaters with 3 empty seats usually? So every car has four wheels, its own computer, its own engine, and so on? In intersections, even with self-driving technology that does not exist currently and is showing no signs of existing any time soon - even with that technology, you’re still threading individual cars through each intersection? It’s not great.

Third. I’ve seen no evidence that the people talking about self-driving cars as a shared or car share-like resource actually believe in it. Remember Musk talking about how public transit is gross because other people breathe on you? Or the tendency of Americans to treat their cars as sacred property, so much that they’ll get in a fistfight with anyone who touches them? Transforming self-driving cars to some sort of public-ish transit requires just as much worldview shifting as actually using public transit, but none of the benefits.

So: if self-driving cars suddenly exist, sure. But right now, there is no baby. There is only bathwater.

1 comments

I totally agree with everything you're saying, there's definitely a long way to go. I am certainly not advocating for going all-in on them any time soon.

I don't see any harm in trying though, or continuing to try and improve them. If we had said that a boom-box sized car phone wasn't useful enough, so why bother trying, where would phones be now?

Besides, you can probably point to any number of people in history that said that humans would never fly, that computers were unwieldy room-sized devices that consumers would never use, that the internet would never be anything more than communication between universities.

It's also just fun to think about the possibilities! :)

I’m totally fine with folks trying - plenty of companies want to build self-driving cars and investors want to buy into those companies, and all that can just sort itself out and win or lose.

The problem is when folks - and I’m not saying just like this thread, but this is a phenomenon in some governance - when folks say that self-driving cars mean that we should reduce public transit or rail investment now, because those problems will be solved, right around the corner, as soon as the cars work.

If/when self-driving cars happen, sure, that’ll be great! People who are skeptical about the technology aren’t going to stop Tesla or Waymo from forging ahead. But car-centric planning does, currently, reduce public transit investment, and self-driving car hype does reduce the political will to build rail.