Classic anti hype cycle. Won't age well.
They have to operate in weather. Rain, snow, dark. They have to adapt to novel conditions.
I put the self-driving car in the same category as the flying car and jet pack. This one just happens to have VC funding.
I don't think we'll see these things for another decade, and when that next decade arrives we'll be saying the same. Another decade.
But no safe full self-driving for a while.
Although I do feel like the money spent in the US so far would've been better spent setting up better public transit/train lines.
There's very little competition because there is not much to compete for, except in longer distance rail and bus.
You also need permission from the city to use public stops, consult routes and times, which is a barrier of entry. Not an insurmountable one and nobody prohibits you from running private stops or garages. You can even negotiate the use of bus lanes or tram lines.
The other barrier is much higher capital investment. Both bus drivers, busses and stops are not cheap, while the margin on a ticket is thin. (Compared to taxi fleet.) But it can be done even encouraging competition - see South Korea for an example of where it worked.
How much are talking? Waymo is 10 billion in the hole. Tesla spent maybe around that much developing autopilot. This is also not public money being spent.
A typical public transit system costs orders of magnitude more than that. Example: California's High Speed Rail project started at 40 billion and projections have already ballooned to 98 billion.
How could you even calculate this? By running massive experiments and seeing if fewer people die over a given period of time driving versus self-driving? Do you really think we can run such an experiment?
It is easier to do it if you consider 1% worst drivers, either older people with long reaction times, distracted drivers, drunk drivers etc. If you have a car that is obeying all traffic rules, it never goes above the speed limit and it has sensors that makes it brake before rear-ending other cars you are already better than the 1% limit. It is a simplistic approach, just giving an example of how 99% is not that hard to reach.
No they don't - a self driving car that knows when it can't drive is perfectly acceptable - not everyone can drive in extreme conditions reliably - doesn't mean they can't drive.
I don't agree, because of the grey area between the car knowing it can't drive and believing it can, but in practice doing so poorly.
That depends on your perspective. I think most of us just wants it to be more reliable than a human driver that's not drunk or sleep deprived. But the liability challenges could mean you'll want more "perfect" reliability.
> They have to operate in weather. Rain, snow, dark. They have to adapt to novel conditions.
Seems like the solutions out there are already getting pretty good at handling the kind of rain, snow and darkness you'd see in most urban/suburban areas. Maybe you'll "never" be able to drive autonomously to your cabin in the woods, but so what?
> I don't think we'll see these things for another decade, and when that next decade arrives we'll be saying the same. Another decade.
What exactly do you mean by "seeing these things"? By some measures they're already here. The ones that are truly self driving are slow, and mostly on static routes. The ones that drive on highways need to be monitored by a human, but they're doing pretty well. I think you'll see these different solutions just slowly become better and better. I'd be very surprised if you couldn't hail a fully autonomous robotaxi in many urban areas in a decade. Maybe it can't drive it anywhere, maybe you can't use it in a snowstorm, but it'll be a case of "good enough".
That is a seriously low bar, and in no way do I think the majority of people feel that driving "better than a drunk" is good enough. In fact, I think it will be the opposite: people will want it to be almost flawless before it's accepted.
>and in no way do I think the majority of people feel that driving "better than a drunk" is good enough
If that the case something can be done such as smart marketing/social media campaign or something like that to change people perception.
> more reliable than a human driver that's not drunk or sleep deprived
emphasis mine.
No thank you. Almost all the humans I share the road with are neither drunk nor sleep deprived. Putting robotic drivers on the road that make daylight driving as dangerous as 1am driving is not my idea of progress. It'd be a bloodbath.
Most of the roads may be clear, but “all” of them never are. Shaded areas are the worst, and can remain iced over for months at a time. Ice, especially intermittent ice, is bad. Slush, deep snow, and packed snow can be miles worse.
Winter driving, as any Californian who has migrated north will tell you, is a dramatically different skill than driving on the roads of California or Arizona. It’s non-deterministic - turning the wheel doesn’t always turn the car. The car will move sideways on the road with no steering input. The speed of the wheels is completely divorced from the speed of the car.
Flying cars and jet packs are limited by physics (specifically energy density) and progress in this area is almost non-existent. We barely got batteries to be vaguely competitive with internal combustion.
Self-driving cars are limited by software and hardware accelerators and this field is making real progress.
I remember a friend saying my kids would never need to learn how to drive. My eldest started her driving lessons last month.
People also posted this about 3D TVs and Cuecats.
Self driving vehicles will save $250+B/year in labor costs in the US alone - this is more impactful than the cotton gin and rivals the impact of refrigeration. 3D TVs and cuecat are 3-6 orders of magnitude smaller impact
Finally, it's worth mentioning that cuecat became QR codes which took off after getting built into cellphone cameras. 3D television didn't take off but 4K did i.e. consumer demand existed for richer TV images just not specifically 3D. Projectors remain niche as well even though they promise to take TV from 80" to 130+" - if ultra short throw (UST) projectors come down in price, I could see them going mainstream, especially in tandem with 4K/8K.