[Theodores]

Very impressive. However these autopilot features are getting subjectively less impressive as time goes on. I want a bit more 'wow'.

Autopilot is not so compelling if you live in Europe with 'analog' roads rather than the mix of well designed, nice and wide 'residential', 'street level' and 'interstate highway' roads that there are in the U.S.A.

Reminds me of Space Shuttle launches, at the start of the programme people would be transfixed for hours waiting to see what would happen, a couple of years later it was 'meh', no need to put the TV on. Even if there was a cool spacewalk some of the magic was lost, awe levels dropped from 10/10 to 7/10. Autopilot is a bit like that, kind of need to see it work going through the streets of Paris, Cairo or Bombay for it to be back in the realm of stunningly amazing.

[antpls]

Living in Paris center, I'm convinced it will be the last place where autonomous cars will be allowed. Roads are smalls, the city plan is old, there are experiments with new e-scooters, people don't respect priorities, etc

It could only work if ALL vehicles were automated and no human driving was allowed. But even then, walking people will exploit the overly-safe vehicles' decisions and just force the priority (which is probably a good thing to give the city back to walking people).

In my opinion, the capital cities won't be the same once autonomous cars are fully developed, there will be probably more public transport, and more car-pool sharing drive services, more escooters, etc. The most interesting long term usage of autonomous driving is on highway and smaller cities than capitals, with less traffic

[wffurr]

Interesting that you consider US highways and streets to be well designed, given that they are objectively substantially more lethal to their users than European roads.

[moogly]

There might be additional factors.

a) training and education - it is way more difficult and takes longer to get a driver's licence in Western Europe than in the US

b) mandatory vehicle inspections - I think this varies by state in the US, but even the strictest state legislation is still less strict than Western European countries when it comes to allowing dangerous vehicles on the roads

c) less regulation for truck drivers' work hours - I think in the past decade, sleep deprivation amongst transportation workers and the results thereof has received increasing exposure in the press, which is good, but awareness only goes so far

[wffurr]

Sure. But highway design, car centric cities, and high-speed suburban arterials are a factor as well.

The deaths per VMT measure is EU-wide, not just Western Europe.

[bane]

https://en.wikipedia.org/wiki/List_of_countries_by_traffic-r...

[wffurr]

Yup, sure looks like Europe is substantially safer than the US in that map.

[RayVR]

In what way are they “objectively substantially more lethal?”

[wffurr]

Double the deaths per VMT.

https://www.newsweek.com/2014/04/04/you-drive-american-24806...

[RayVR]

I think attributing that to the design of the roads is kinda silly. We have very low standards when it comes to acquiring a license to operate a vehicle.

[wffurr]

I think attributing it to any one cause is kind of silly. It's a fact that American cities are much more car centric with higher speeds and fewer areas with traffic calming measures, all of which make the roads much safer.

Daily highway driving is much more common for commutes and errands, suburbs have dangerous arterial street designs.

I'm sure there's more to it, such as licensing and vehicle safety measures, but American highway design (and making local roads into mini highways) is a part of it.

Also the EU measure is EU-wide, and not all countries in the EU are as strict as say Germany for licensing.

[dkoston]

You’re going to be disappointed for quite some time then. Most of the improvements to autopilot in the near future will be refinement of the cars reactions to sensor data and not new public facing features. They are closing in on the line where regulation does not allow for further “self driving” festures. Having the car change lanes itself is “self driving” and that’s why the driver must initiate the lane change. The same argument can be made about stopping at stop signs and red lights which is the next killer iteration of driver assist. Hopefully there will be enough data collected and good decisions made to allow for iterative improvements to driver assist features which would improve overall traffic safety.

[danielmg]

My 2018 Mercedes E class can do lane changes - I just have to put the signal on and it does the rest. They call it "Drive Pilot". It seems to be just as advanced as Tesla but much less "marketed".

[dkoston]

Yup, Audi offers similar driver assists with lane changes(https://www.audi-mediacenter.com/en/technology-lexicon-7180/...), Land Rover has lane keeping (https://www.landrover.com/ownership/incontrol/driver-assista...), so do Honda/Acura (https://www.acura.com/rlx/modals/lane-keeping-assist-system-...) and many more.

I agree that Tesla overhypes their system the most and that contributes to consumer confusion. They use the classic Silicon Valley tactic of making their features which are comparable to competitors seem light years ahead through marketing. On one hand, it sucks for consumers, on the other hand, this type of advertising is rampant and they are playing the game.

Tesla’s over promising does erode trust. There’s only so many times you can promise stuff that’s obviously overhyped or years out (FSD) before consumers will stop believing the hype.

What’s interesting about Tesla is that almost all of their controversy is self inflicted by Elon’s comments.

What’s not clear to me is if this is intentional as he wants to stay in the news cycle or if it’s driven by his ego.