As someone who spent $8K on Enhanced Autopilot 15 months ago, I'd say that pretty much everyone has better autonamous driving. My brother-in-law got a new Honda Accord a few months ago, and says on the highway he can just take his hands off the wheel and let it do it's thing.

My fancy-schmancy Tesla Models S meanwhile can do an ok job of following lanes as long as you keep your hands on the wheel and pay attention to ensure that whenever it doesn't detect your hands on the wheel that you give it a little tug.

I love my Model S, but they really bit off more than they can chew in the "writing our own software" department.

Do you have 2018.10.4 or above? It drove me through 220 miles of rain including torrential downpours yesterday with ease. Up until that update, I’d agree that AP2 was sub par but the new update has dramatically improved EAP.

> It drove me through 220 miles of rain including torrential downpours yesterday

This sounds terrifying. Why did you trust it enough to experiment with your life?

And the lives of everybody else on that road.

I just got 2018.10.4 a week or less ago. I haven't put many miles on it yet, but it does seem like a big improvement in the small testing I've done, as far as the driving.

> My brother-in-law got a new Honda Accord a few months ago, and says on the highway he can just take his hands off the wheel and let it do it's thing.

I have the new Accord. It's not really autonomous. Adaptive cruise control and auto-braking aren't bad, but the lane keep mostly just jiggles the steering wheel to notify me when I depart the lines, and all the time it mistakes construction and seams in the pavement as lane lines.

Compared to the current, non-MobilEye Teslas, the Leaf, the A6, the CT6. Even the old Model S has better autonomous driving than the current one.

Oh, and Google. By very, very far.

Tesla hasn't even caught up to its original Mobileye-powered Autopilot.

Do any other autopilot systems from legacy automakers improve with over the air software updates? Do they revise their algorithms with fleet data?

I can't get Waymo/Google autopilot in a vehicle I can buy. I can get Tesla's autopilot.

The latest Tesla autopilot software update (2018.10.4) has shown drastic improvements (30+ Youtube videos from owners in thread): https://www.reddit.com/r/teslamotors/comments/84xt43/collect...

> Do any other autopilot systems from legacy automakers improve with over the air software updates?

Who cares? I'm not buying a luxury vehicle on promises of what it might do given enough time.

To each their own of course. I'd have no qualms about buying my Model S again.

I'm unwilling to buy a car stuck in time in it's model year, or that must wait years for a software upgrade. Then again, I'm happy with my iPhone, bugs and all; it'll get updates too, and works just fine between iOS updates.

> I'm unwilling to buy a car stuck in time in it's model year

You know I used to think this way, then all my computing devices started changing often for the worse with every update.

For me a stable device that doesn't change month-to-month is a feature, not a disadvantage. Give me the LTS branch of the car world.

I can appreciate that perspective. I want to live at the bleeding edge, and don't mind the inconveniences or regressions. I can't take my money with me when I die, so after savings and basic needs are met, I want to live in the future (even if the products are still in beta!).

I can't get that sort of experience from stodgy, risk adverse legacy enterprises. I want to support enterprises that want to succeed or die.

EDIT: Edit after your edit, keep the LTS builds, give me nightly! Tests pass? Ship it!

Please don't beta test your products on the same roads I'm using with my stable branch ;)

> Tests pass? Ship it!

If there are bugs, more tests! And always, always, always degrade gracefully.

> I want to live in the future (even if the products are still in beta!)

Now we're getting in to why traditional automakers are 'stodgy' and 'risk averse'.

I do agree that there is a lot that can be improved in cars, but a ship software at a million miles an hour approach seems like the kind of practice that will end up externalizing the costs to the rest of us.

Edit: I'm gonna leave this up, but it doesn't really reflect the friendly tone I'm trying to put across - I've enjoyed this chat.

This is the kind of self-centered, callous mindset that runs over innocent pedestrians.

It’s a car, not a web site. People other than the drivers’ lives are at risk when you drive on public roads.

I’m the one who is responsible for the vehicle at all times, not an Uber safety driver who isn’t paying attention, and I take that responsibility seriously.

Can you be isolated to a testing serv^H^H^H^H roads?

Maybe you just don’t want the pain of feeling like you wasted 100k.

I only spent a bit more than half that on a CPO Model S, and I'd do it again in a heartbeat. If you have the means, I highly recommend it. By far, my favorite car having owned several Corvettes, an SLK 350, a luxury pickup truck, and a Subaru STI.

Being a "leap forward" still puts it behind others. But if your criteria for quality is being able to hope that one day you'll get an update, well, that still puts you behind Cadillac and possibly others.

You can continue to hope and wait for Tesla to deliver. So far, it hasn't.

Ironically Autopilot 2 is better than Autopilot 3, so older Teslas would be one car.

Given the accidents, you might as well drive a regular car and take your hands off the wheels and you'd roughly replicate the quality of their "autonomous diving".

I tried the Audi A6's self driving feature and it was fantastic.