I think the data advantage is limited.

I used to work in high-frequency equity trading. We would run simulations against tick-by-tick data from real markets. The simulations were of very questionable value. As soon as you trade once in your simulation, from that point onwards the whole run is contaminated. Your trade, if it had actually happened, would have been seen by other participants, who would have changed their strategy, often in complex and unpredictable ways, and the timelines would have diverged rapidly and irreversibly from that point.

Data recorded from driving situations is useful for one thing - testing out split-second decision making - stop or accelerate? Steer left or right? This is a limited use case. Most situations don't have critical decisions like this. The ones that do, it's often hard to judge what would have been the right answer. Even accidents - where the post facto desired outcome is "don't have an accident", unfold over several seconds, involving several stages of actors responding to each others' actions. No one can be sure how the driver should have approached the very first interaction - the only one for which the data is meaningful.

Actual road testing, and meaningful simulation which isn't just replaying what actually happened is much more useful.

> Waymo

Even they admitted fully autonomous vehicles most likely won't be a thing any time soon (maybe ever if we keep our current infrastructure)

https://www.cnet.com/news/alphabet-google-waymo-ceo-john-kra...

I feel like this was misread. The statement appears to say that he doesn't think they will completely displace manual driving of cars for considerable time. And that SAE level 5 arguably expects more than we expect from human drivers (many of us would not be safe driving in whiteout conditions on a windy mountain road).

I did think it was imminent at one point, but the fact the delivery date keeps pushing out shows its far from complete, and is only usable in controlled circumstances. I use it in traffic jams, or motorway driving, and that's it, anywhere else and it's just asking for trouble. It's useful in those limited cases, but it's a shame how many false promises have been issued, I've a lot of respect for what Elon has done, but he's really oversold this

I actually have watched a number of his videos, and while it is considerably better than the Tesla self driving, it's not something I'd currently pay to experience vs driving myself (or vs an Uber at a similar price point). The driving style is still too awkward and, to me, discomfiting.

Ah, to me it looks pretty good. And there's reason to believe it'd be better on freeways.

I'd definitely pay a few bucks per hour for my car to drive me to Southern California while I read a book.

What I don't understand is: why don't you take a train instead?

I have never been to US, but whenever I hear about self driving cars I think: why don't people just use public transport if they don't want to drive?

The US has a great rail system if you're a container or other freight that is easily put in a rail car, and you don't mind travelling at low speed and making a fair number of stops and not having any specific date and time of arrival.

It has a terrible rail system if you're a passenger. Connectivity is weak, and even if you have a direct route, almost all the rails are owned by freight lines, so freight traffic gets priority and schedules are rarely met. If you have to make a connection, good luck. There's some exceptional routes in the north eastern US, where there's a lot of rail and some of it is passenger exclusive.

The question is more likely to be flying vs driving than take a train vs driving, because a train will nearly always take longer and be less convenient.

Strangely, there are no direct trains from northern CA to southern CA. There is a line from Sacramento down into Santa Clara, and a line from San Luis Obispo down to San Diego.

But there's no direct way to get from SF/SJ to LA/OC via train. There are buses, but they are considerably less enjoyable than the Pacific Surfliner line!

It’s too much of a hassle to use public transportation.

With your own car you can bring as much stuff as you want, stop where you want, leave when you want, go wherever you want at your destination, bring whoever you want, listen to what you want on the radio, stop where you want for side trips/tourist things along the way etc.

Like for instance with my own car instead of a shared train, I can bring my golf clubs, my dog, and my drone with me, and call my friend on the way out of town to see if he’s up for a spontaneous trip to San Diego to see what fun we could get up to there for the weekend. That’s not really possible on a train/airplane.

It’s just a totally different experience than arriving to a train station with some limited set of your things, possibly being late, then arriving somewhere else again without a car, and having to return again by doing all of this in reverse.

It also means spending time forced to socialize with people you don’t know, and almost definitely in a seat that is less comfortable than the one in your $100k luxury car.

The vast majority of car trips are nothing like this though. Why don't we build more trains to take people on their usual short commutes and save cars for the occasional fun trips like this?

Because of the last mile problem.

US population density is typically low, even in many cities.

This means that any reasonable spending on buses and train routes ends up with relatively infrequent service to places not too close to you.

This is self-reinforcing, too. Everyone taking cars for all trips means you need lots of space for parking and roadways, which spreads things out and makes walking and transit less practical.

> With your own car you can bring as much stuff as you want, stop where you want, leave when you want, go wherever you want at your destination, bring whoever you want, listen to what you want on the radio, stop where you want for side trips/tourist things along the way etc.

Also, you can get blocked in a traffic jam when you don't want.

Urban geography: In 1947, the US started to build very low density, cheap, mass produced suburbs. Along with this, most American cities have very little mixed use zoning, and large parking requirements, both of which increase the typical distance a person would have to travel to live a normal life, and make each theoretical train station/bus stop less economically sensible. When in Spain you might have many tens of thousands of residents within a 1 km radius of a subway line, you will not not get one thousand in a deep suburb. Therefore, for the same total population, a US public transit system (outside of NYC, and maybe Chicago) is going to be far worse than the equivalent on most European cities.

This is self perpetuating: Changes in urban planning that would allow most homes in the US to be good residences without a car would involve not just demolishing most cities and starting over. This isn't even just about demolishing suburbs: Even downtowns are built in ways that optimized for the car. Go to your favorite mapping app, and look at, say, St Louis. You will find many buildings are just parking lots, very few storefronts, very wide streets designed to be driven at 60 km/h, and with very few pedestrian crossings... All the infrastructure that makes it easy for most workers to live 20-30 of wherever they work just makes building a useful public transport system unfeasible for a city, and even more so when most voters would be unable to use it.

The same problem is there in small cities, and even villages. A small town in Italy that lives from summer tourism is going to be easy to deal with even without using public transport, and just walking. The American equivalent is unmanageable without a car.

Those that want to avoid driving in the US end up either paying people to drive for them all the time, or live in the pedestrian friendly parts of a handful of cities. And if you haven't visited the US, you probably also haven't seen how expensive is to live in those areas. So, if anything, it's much easier to live without a car if you are a pretty wealthy tech worker than if you are a server at a restaurant, or work at a retailer. The less your job pays, the harder it is to not drive a car.

Imagine a couple of a bank teller and an insurance agent. If they need to move to Madrid, it's very likely that they will find living arrangements where it's easier to go to work via public transport, as most places do. If you do the equivalent in Houston TX, it's almost guaranteed one of them will have to drive to live a comfortable life, and most likely both.

The US is spread out to an extent that people in Europe have a difficult time imagining. This tends to make people car dependent, which in turn impedes public transit even where it'd work well.

The US has good train infrastructure for freight (far better than Europe), but rail service is not great. There's no good train route from the SF Bay Area to Southern California. But even if there were, getting around in the LA area practically requires a car, so you'll be renting one at the end of a train journey anyways.

So, the options to get my family down there look like:

* $100 roundtrip flight * 5 of us = $500 plus the cost of renting a car. Annoying airport logistics and luggage, and a few hours of time saved

* ~$300 of wear on vehicle and gas round trip, plus having to drive for 6 hours.

A good train would be something inbetween-- perhaps $30-40/person, but still needing to rent a car. Somewhat easier logistics, but still annoying luggage, etc. Time between that of a flight, etc.

But a self-driving car would take the worst suck out of the driving trip.

Obligatory answer from someone that doesn't meet one of the requirements - I'm in the UK so can answer from that perspective.

Going to the city near me, I have these options to take two adults and a child:

1. Train - 5 minute walk, 1 hour travelling including a change. Trains run once per hour, costs £16 while my son is under 5.

2. Bus - 25 minute walk (at adult speed), 45 minute travelling each way. Runs every 20 minutes I think, costs ~£8.

3. Drive - 25 minutes travelling each way, no walk. Costs ~£1 in petrol + £2 tolls. Technically costs would be higher if you include wear & tear on the car itself but not by a great deal.

This is probably the least fair comparison to the car as well. Most other places nearby have the same major issues with travelling, or worse. A trip to the beach would be a 10 minute drive with no tolls but the same 25 minute walk to the bus stop and then a slow bus.

Added to this the amount of stuff we may have on us (pram, bag of various things for changes, snacks, lunch, etc). If there's more of us, or once our kid grows up more the costs go up for everything other than the car.