Where do you think Tesla's FSD will be 12 months from now (April 2020)?

[EVNow]

So I think you're saying you want level 2 city NOA? I am not sure why that is a particularly useful feature, quite aside from the possible safety issues.

It would be very useful - if it starts stopping at traffic lights/stop lights, makes turns etc. I monitor it closely even now - and - I've to keep disengaging/engaging. So, much better than now.

It seems like maybe @Daniel in SD was saying that at least Level 3 NOA on the freeways is a more tractable problem (I'm not so sure). I'm not sure why he wishes for L3 NOA on the highway to be honest. It seems incredibly frightening at this stage.

Same here - I'm happier going around on AP in slow speed suburbs, than at 65 mph on freeways.

Personally, as an investor, I'd like Tesla to focus on safety and making cars that people want. This FSD stuff is kind of silly.

If Tesla wasn't talking about FSD - everyone would be asking Musk about it and saying they will not be able to compete in the future when all the others have FSD and Tesla doesn't.

As Musk said the future depends on Battery & FSD. They already spend a lot of time & money designing safe cars and adding safety features.

 

[Octo]

Its no different from Waymo/Cruise selling FSD to their investors. Tesla isn't saying they already have FSD working, either.

It’s different in various ways. For example, Waymo doesn’t sell you a car today with “FSD” as a real cash option.
They don’t sell you a car with the claim “it will find you in the parking lot. REALLY!” ($6000 extra).

 

[AlanSubie4Life]

You have to shoot for the moon.

You can't score without goals.

If Tesla wasn't talking about FSD - everyone would be asking Musk about it and saying they will not be able to compete in the future when all the others have FSD and Tesla doesn't.

This is true.

As Musk said the future depends on Battery & FSD. They already spend a lot of time & money designing safe cars and adding safety features.

On the safety front, things have only just started; we have no idea whether Tesla's accident rates are lower than other comparable cars - for sure the difference is small if it exists (way less than a factor of 10). If a car is capable of FSD, long before that, it will be capable of avoiding nearly all accidents. I'm proposing that first, you have to win the battle with the public (and the technology of course). Roll out safety features showing preternatural accident avoidance. You have to have those anyway! It's much safer to roll them out initially as background safety interventions, rather than in-your-face "the computer is driving" features. When the public sees that computers are better at driving than humans, as they fix human mistakes (there will be lots of dashcam footage), you start to win the perception battle. It may not be what Tesla drivers want (or exactly what they paid for with FSD), but it's what the public trust likely requires.

It also buys time to actually get the reliability to where it needs to be, and saves lives in the process. Remember, once you've added 3 9s (or whatever) to human driving with machine assistance, you'll also need to add another 9 (or perhaps 2) to the FSD safety requirement to get to an acceptable level relative to state-of-the-art human (computer-assisted) driving.

Once the safety levels come up, that's actually a demonstration that FSD is getting close to ready - and there will be massive amounts of data at that point to figure out where issues may still exist.

 

[Daniel in SD]

Its no different from Waymo/Cruise selling FSD to their investors. Tesla isn't saying they already have FSD working, either.

What do mean? They just need regulatory approval!

 

[EVNow]

What do mean? They just need regulatory approval!
View attachment 435766

Most people already think regulators will not approve until FSD is very safe (probably true in many parts of the world, not called Florida). So, it covers both

 

[EVNow]

If a car is capable of FSD, long before that, it will be capable of avoiding nearly all accidents.

Not really - the car can drive defensively, it can't force the driver to drive defensively.

Anyway, this line of discussion is OT.

 

[Daniel in SD]

Most people already think regulators will not approve until FSD is very safe (probably true in many parts of the world, not called Florida). So, it covers both

I just wish Tesla would shut up about regulators until they actually have something they think is ready for approval.

In other news I'm building a cold fusion generator in my garage and after I get regulatory approval I'll be able to charge my Model 3 for free

 

[Daniel in SD]

Not really - the car can drive defensively, it can't force the driver to drive defensively.

I think it could but people wouldn't like it very much. haha.

 

[EVNow]

I think it could but people wouldn't like it very much. haha.

Yes - just imagine the car enforcing the 3 second rule for following other cars.

Or better, not starting the car if it detects the driver is drunk

ps : I think regulators aught to make some of these mandatory in cars, including the simple rule of automatically stopping the car before hitting anything. All the technologies are there and quite cheap.

 

[AlanSubie4Life]

Not really - the car can drive defensively, it can't force the driver to drive defensively.

Anyway, this line of discussion is OT.

Unless people forget how to drive, presumably they will continue to drive approximately the same way.

Regardless of people not driving properly or defensively, I would say that there is definitely room for actual accident avoidance rather than mitigation. FSD will have to be able to avoid accidents (ones that occur when driving defensively) in order to provide a significant enough safety advantage.

I don’t think this is off-topic at all - it speaks to the trajectory of FSD and how Tesla will roll it out. My contention is that Tesla will implement lots of new safety enhancements and accident avoidance far before being “feature complete.” I just think it’s a much easier (this does not mean easy) problem, and I think their hardware will have substantial ability to provide such enhancements. In addition, it will come out as a natural consequence of actual FSD feature development. The interventions are one of the only ways for them to find out how well FSD might work if it actually progresses past level 2.

I suspect that Tesla’s analysis of what will happen if they just roll out city FSD/NoA will indicate serious safety issues with driver attentiveness, so instead, rolling out continued safety improvements will be a way of capitalizing on their development (even if they won’t be able to recognize revenue for their FSD sales with this approach, it will help reputation and sales).

 

[EVNow]

I don’t think this is off-topic at all - it speaks to the trajectory of FSD and how Tesla will roll it out. My contention is that Tesla will implement lots of new safety enhancements and accident avoidance far before being “feature complete.”

That roll out plan is definitely news to Musk

 

[AlanSubie4Life]

That roll out plan is definitely news to Musk

Sure. Would not be the first time!

 

[Daniel in SD]

That roll out plan is definitely news to Musk

They just rolled out Emergency Lane Departure Assist and they still haven't rolled out Enhanced Summon. Safety features are easier to implement and have real world benefits today.

 

[EVNow]

They just rolled out Emergency Lane Departure Assist and they still haven't rolled out Enhanced Summon. Safety features are easier to implement and have real world benefits today.

From an investor POV, that does nothing to the stock price.

The point is - that is definitely not what Tesla's aim is. May be in a sense it is - Musk thinks FSD will make driving safer (you hear the same thing from Lex Fridman, for eg).

Anyway as I said, this is OT. I think OP should start a new thread, if he thinks that is how the roll out would be.

 

[dgatwood]

I would like a youtube channel with videos showing people driving for miles and miles, like an entire 45 min-1 hr commute, with AP on, without disengagements. I get that it would be boring to watch, but don't want it sped up, need sound and everything. I have seen one TMC poster post his entire commute (I think when he got NOAP without confirmations) that I need to find again. Seeing how others are using AP would really be helpful to me.

I drove CA-17 southbound from Los Gatos to Scotts Valley on Sunday (or was it last Thursday?) without any disengagements. I was very tempted at a couple of spots, though, because it still likes to hug the concrete barriers a little too tightly. Northbound is still not there yet.

 

[boonedocks]

You have to shoot for the moon. That is how the once impossible becomes possible. If we didn't set far out goals and miss a few times at first, we'd never achieve anything as a human race.

Its no different from Waymo/Cruise selling FSD to their investors. Tesla isn't saying they already have FSD working, either.

I guess one huge glaring difference is, How many of those FSD cars / features did they actually market and sell go people (general public buyers)? Three, Two, EVEN One?

 

[cucubits]

Wonder why the poll has been closed since this is an active thread on the topic.

Did the OP not like the results?

 

[iqless]

Who do you guys think is the funny person who thinks Tesla will have full L5 autonomy next April

 

[diplomat33]

Who do you guys think is the funny person who thinks Tesla will have full L5 autonomy next April

That would be @electronblue:

Q2/2020 then. On the 22nd Elon Musk said Level 5 no geofence is feature complete (Q4/2019) and should be live somewhere in the U.S. (Q2/2020), so I picked that — Tesla’s FSD is L5 autonomous! — solely based on Elon Musk’s information who should know best.

 

[iqless]

Aww.