Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

Tesla, TSLA & the Investment World: the Perpetual Investors' Roundtable

This site may earn commission on affiliate links.
Or a section of NN trained to detect the aspect change of an approaching pothole edge vs the normal planar surface. Vision might work better against water filled potholes also...

That'd take some pretty extreme camera resolution until the pothole is just a fraction of a second away from you. E.g.: if you're driving 35m/s and the lip of the pothole is 1cm, and the camera is 50cm above the road, then when you're one second away from it, the bottom of the pothole is 0,000004% further away from the camera than the top of the pothole. There will be virtually no parallax. Or are you talking about recognizing "pothole shapes"? IMHO you're going to get a ton of false positives with that.
 
Last edited:
  • Like
Reactions: humbaba
I have an issue with the notion that there is a single "FSD". Self-driving can always get better. Are we going to say that just because it's possible to make a system better in a future generation of the vehicle, then the previous generation can't be called self-driving? If that's the case, then nothing will ever be able to be called self-driving.

Hardware will keep getting better. We need to accept this, not insist on locking us into one particular generation and cursing Tesla if they ever introduce something even better. The software will always do as well as it can on any particular hardware set.
Of course it will get better. And don't let perfect be the enemy of good. But when the problem of potholes was pointed out you give conjecture about possible future hardware. That's great for the future, but potholes exist in the present. As does a need to hear car horns, sirens, etc.

FSD, outside of limited use cases, is not just around the corner.
 
  • Like
Reactions: Mader Levap
Of course it will get better. And don't let perfect be the enemy of good. But when the problem of potholes was pointed out you give conjecture about possible future hardware. That's great for the future, but potholes exist in the present. As does a need to hear car horns, sirens, etc.

FSD, outside of limited use cases, is not just around the corner.

Hearing a horn/siren/etc isn’t necessary. Those sounds exist to bring human attention to look in a particular direction, see what the issue is, and take the correct action. The networks in Tesla’s case are already paying attention to everything in the 360 degree view around the car. They can already see the car that’s honking or the emergency vehicle with its siren going.
 
  • Disagree
Reactions: Mader Levap
Hearing a horn/siren/etc isn’t necessary. Those sounds exist to bring human attention to look in a particular direction, see what the issue is, and take the correct action. The networks in Tesla’s case are already paying attention to everything in the 360 degree view around the car. They can already see the car that’s honking or the emergency vehicle with its siren going.
Right, because there is always a good line of sight to where an emergency vehicle is. Here in the real world there are large vehicles, buildings, etc. Similarly, it can "see" a car (at least, as long as it is moving), but it can't see that it is honking its horn.

edited to add:

More particularly, in reality, hearing a siren often doesn't tell you where it is. But it is a caution to slow down and be observant. Sure, emergency vehicles will virtually always cautiously negotiate intersections and so an FSD car may get by with only visual perception. And it would likely slow down if other traffic did, but what about if there isn't traffic in front of it to slow it down as it hits the intersection and green light just as the emergency vehicle enters from the side?
 
Anecdotal data point:

Since I reserved my Model 3 I started telling people that I am getting an American car, a claim that is _every_ time met with disbelief (maybe because I am known for driving Audi for 10+ years).

When I then explain that I am buying a Tesla, I get a: "But that's different".
So I think Teslas will (continue) to sell just fine here.

Anecdotal data point too:

Spent about 20min “selling” Tesla and the M3 specifically to a VW sales guy dealing with a leasing return of a company car. He told me he drives the eGolf (PHEV) and complained bitterly about the fact that his 45km all-electric range was down to 25km in winter, which meant he couldn’t make the commute to work on electric only....

He agreed that VW was years away from anything to compete with the M3. I suggested he just go for a test ride - that usually answers the why? question immediately.
 
final_5c61d90aa2a94f0014d2ed68_470243.jpg
 
Good point, EVNow. Nevertheless without spreading advertising dollars, the Tesla PR team is in a weak position for effectively influencing the media.
If strategically placing advertising is what it takes to make the FUD go away, unfortunately that is what Tesla will have to do :( It is probably cheap in magazines, newspapers. Instead of cars, they can advertise the brand or some do some public interest ads (like about climate change etc). Probably wine & dine reporters. They have to play the game ...
 
Of course it will get better. And don't let perfect be the enemy of good. But when the problem of potholes was pointed out you give conjecture about possible future hardware. That's great for the future, but potholes exist in the present. As does a need to hear car horns, sirens, etc.

FSD, outside of limited use cases, is not just around the corner.

Are we pretending that human drivers don't hit potholes every day?
 
Well, let's hope it works out better for the guy than the SunEdison "thingy" did...
Odds are he'll nail one eventually, right?

View attachment 376358

My god, you're right. He recommended another company's stock four years ago and it tanked! I'm closing all of my positions immediately -- thank you!!!!
 
Right, because there is always a good line of sight to where an emergency vehicle is. Here in the real world there are large vehicles, buildings, etc. Similarly, it can "see" a car (at least, as long as it is moving), but it can't see that it is honking its horn.

edited to add:

More particularly, in reality, hearing a siren often doesn't tell you where it is. But it is a caution to slow down and be observant. Sure, emergency vehicles will virtually always cautiously negotiate intersections and so an FSD car may get by with only visual perception. And it would likely slow down if other traffic did, but what about if there isn't traffic in front of it to slow it down as it hits the intersection and green light just as the emergency vehicle enters from the side?

If it can’t see it, neither can a human driver. And emergency vehicles don’t tend to just blast full speed through red lights at a blind corner(siren or no, a human is likely to crash into them if they do). In such conditions, they have the ability to change the light to green before they get there.

For car horns, if you can see what the car is doing, hearing it’s horn is pointless(ditto if you can’t). All a horn does is say “look over here”. The networks are already looking over there and tracking that car, along with all other cars.