Separate names with a comma.
Discussion in 'Model 3' started by osubuckeyes88, Jun 27, 2018.
What are your thoughts/opinions?
Question is worded weirdly.
You understand that the price is still $3,000 during vehicle purchase?
If the question is "Now that Tesla is increasing the post-purchase price difference for FSB from $1000 to $2000, will you be including FSD in your initial purchase?"
Very unlikely. The feature they roll out in August would have to be blow-away amazing by itself, because I still don't have confidence Tesla (or anyone on the planet) is anywhere near [Level 5] FSD.
I was on the fence a little because none of the features have been released. I can't ignore that $2000 penalty for not buying it upfront.
In addition, if a computing hardware upgrade is needed I'd expect to get it for free.
I'm betting FSD never makes it the current M3. I don't think it has enough sensor or processor redundancy to satisfy regulators.
You're supposing regulators already have regulations. It just needs to have enough redundancy to pull over to the side of the road with the hazards on. It's never going to be able to complete the journey if the sensor fails.
Once some self-driving features do start getting released, eventually leading to L4 self-driving in a few years, the "FSD" package could easily be worth $5000 or $6000. So getting it now for $3000 could be a real bargain.
How many eyes do you have? How many are required for driving?
How many processors do you have?
I'll agree with you, most of the people on the road shouldn't be driving. Not enough redundancy and a really bad track record.
On the other hand, the Tesla has stereoscopic view, with redundancy, 360 view, with the addition of distance sensors and radar. Oh, and the cameras have more resolution than the human eye.
The human is definitely out-classed here.
Have a source on that. From my understanding, human eye have some ridiculously high resolution, which is why we can see pixels on 1080p monitors from not too far away, whereas the tesla cameras aren't super high res
The only way I could justify paying a total of $10K for FSD is if I was going to use the vehicle commercially to make money. The $5K I paid for EAP was hard to justify but if it helps me be safer I could rationalize it. It isn't there yet and FSD is a long way off unless they have a completely different software sweet than what we currently see implemented. The visual recognition isn't there nor is the fusion between cameras and radar. They also need rear radar units that are missing too. Given the number of bugs we see in each release, it is apparent to me that the testing sweet they have is inadequate or they don't test very much. It appears they often break things that were working wile adding some new functionality. I hope the coding isn't that sloppy but from the outside it sure looks that way.
Human eye resolution is ridiculously low in the fovea (which you use to focus on something). The rapid movements of your eye compensate.
Cameras don't have this issue and can focus on multiple areas at once.
Right, but the middle is very high resolution and our eyes also twitch a little which can help increase the resolution
Already purchased for $3,000. Not an option in the poll.
Something I found online:
According to Dr. Roger Clark, a digital and film imaging professional, who is also involved in several outer space imaging NASA projects, the resolution of the human eye is approximately 576 megapixels—which is a conservative estimate because this only pegs the viewing angle at 120 degrees, and the human eye actually spans a bigger field of view of 180 degrees.
Depends if you consider 0.2 megapixels high resolution (200,000 cones in fovea.)
we discuss it here:
AP2.0 Cameras: Capabilities and Limitations?
I voted no, but will purchase it , after it is proven to work, perhaps in 5 years.
That's a huge issue with it, a dearth of well worked out regulation infrastructure.
This is a bad, bad comparison.
1) I've got a whole set of tools and strategies to address my "sensor issues".
2) Autonomous driving has to be better than at least 90th percentile of humans, maybe even higher than that, to make sense.
The retina contains two types of photoreceptors, rods and cones. The rods are more numerous, some 120 million, and are more sensitive than the cones. However, they are not sensitive to color. The 6 to 7 million cones provide the eye's color sensitivity and they are much more concentrated in the central yellow spot known as the macula. In the center of that region is the " fovea centralis ", a 0.3 mm diameter rod-free area with very thin, densely packed cones.
Keep in mind that rod cells provide an insane amount of information and cone cells are just for color
Mostly agree, though not so sure about camera vs eye resolution. Maybe for people who need glasses
This argument gets floated regularly, and it's still wrong. What the human misses in sensors, it makes up for in an extremely powerful processor, trained for decades, and that is many decades ahead of anything that we'll be able to build.
The current sensor suite is vulnerable to something as simple as a splotch of mud or ice over the forward cameras. You will never have L5 without a lot more sensing redundancy than what we have now. Industrial automation systems that protect human life legally require redundancy in the mechanical and electrical systems - so that if any part fails, life is not risked. I can't see L5 cars requiring less.