Humans are bad at driving. I do not want an automated system to be a substitute, I want it to be vastly superior and safer. 40,000 people die each year in car accidents in the United States. I expect an autonomous system to be ZERO deaths a year. Getting there will absolutely require more robust systems than what humans come equipped with.
back
1 comments
Humans rely on two eyes and only look one direction at a time. Even the current Tesla Vision sees much more at once than a human is capable of.
Also I agree we should aim for zero deaths. But any improvement is better than no improvement.
There's plenty of videos on youtube of what a little fog does to tesla vs any other radar auto-brake system.
My favorite is where they painted a cardboard wall to match the road, Coyote and Roadrunner style and Tesla was the only car to smash into it.
Put a radar/lidar already into the Tesla!
Why not make it an amazing car, just because Elon is a narccisistic/sociopatic asshole?
I swear Elon is his own greatest enemy.
I think he cut the sensors because he's so used to Muntzing — save money by cutting stuff until something fails and then putting it back — being a winning strategy.
That Musk is surrounded by sycophants means that Muntzing is now not going to work in general, because so many people will tell him that everything's fine even when it isn't. Only SpaceX can get past that, because RUDs are impossible to miss.
I understand you don’t like Musk. But regardless of that -
vision vs sensors. Isn’t it about shaving costs to make a mass produced car that’s attainable by more people due to lower costs and easier manufacturing? When considering engineering, manufacturing, and cost only - I understand why it is why it is right now. Even the current Tesla Vision FSD today is amazing. Agree that I’d love to see sensors in the future as long as it doesn’t price me out.
Musk is complicated. I recognise he has done interesting things, my personal feelings to not detract from that. Even my very claim that he is used to Muntzing being a winning strategy is because (as I understand it) it did genuinely help to make the early Tesla models into products that were simultaneously affordable and not loss-making.
As for price, LIDAR has been built into high-end smartphones since at least the iPhone 12 Pro — 4.5 years ago. When LIDAR was new it made sense to avoid it, but I don't think that argument has been viable for a while now.
There's also the loss of ultrasonic sensors, which are cheap enough to be in budget teach-kids-electronics kits.