Please register to participate in our discussions with 2 million other members - it's free and quick! Some forums can only be seen by registered members. After you create your account, you'll be able to customize options and access all our 15,000 new posts/day with fewer ads.
So FSD running over a dummy is fine? I’m lost because the fully self driving software should do something to avoid an object in its path shouldnt it? Or no, running it over is just fine if it can tell the object isn’t alive?
Ideally yes. However you also don't want it to stop for you don't care about like plastic bags or aberrations. Too many false positives make it unusable.
These are tuneable models. For an L2 system, you want to tune it to be cautious but not braking for every single item.
A L4/L5 system will probably be tuned to stop for more objects.
I will say AP and FSD (not accepted to beta) has been a major disappointment for me, but I've been spoiled by really good non mainstream ADAS. FSD beta looks promising but I ain't driving my 500hp like a grandma on the hopes I get accepted, I tried for a week and couldn't get close to a good safety score.
Ideally yes. However you also don't want it to stop for you don't care about like plastic bags or aberrations. Too many false positives make it unusable.
These are tuneable models. For an L2 system, you want to tune it to be cautious but not braking for every single item.
A L4/L5 system will probably be tuned to stop for more objects.
I’m pretty sure it’s not designed to run over dummies that simply does not make sense, we aren’t talking about plastic bags. Do you have any information from Tesla that says it’s designed to run over child sized objects ?
I’m pretty sure it’s not designed to run over dummies that simply does not make sense, we aren’t talking about plastic bags. Do you have any information from Tesla that says it’s designed to run over child sized objects ?
it is designed to only stop for objects it classifies into a certain object set.
Regardless, the issue from an investor's viewpoint is whether Tesla's sensor / compute stack is sufficient to detect small static objects. That is not clear yet to me.
it is designed to only stop for objects it classifies into a certain object set.
Regardless, the issue from an investor's viewpoint is whether Tesla's sensor / compute stack is sufficient to detect small static objects. That is not clear yet to me.
I’m not sure how you can have the first response and then say it’s not clear to you if it can sufficiently detect small objects. It’s clear it’s not sufficient if the car has no ability to avoid hitting a dummy
I’m not sure how you can have the first response and then say it’s not clear to you if it can sufficiently detect small objects. It’s clear it’s not sufficient if the car has no ability to avoid hitting a dummy
No no, 1) it may detect the object. But if 2) it does not label it as a critical object, it will still let the car run into it.
We know the #2 condition exists, therefore we cannot tell if the issue was #1 or #2.
From an investment standpoint, I am focused on #1.
Not surprising that most don't even understand how Tesla's FSD software operates.
It creates a depth map that can find objects in the driving space. It also uses an object classifier to label objects as vulnerable users, or other common obstacles.
The models can indeed not classify the dummy as a human nor any other object to avoid, and therefore proceed to run it over, and the models would be working correctly.
That's not to say that FSD can't screw up. It is an L2 system as you assuredly must know. Let me know what other L2 systems can drive city streets and detected almost all vulnerable road users. Glass is 90% full, not 10% empty.
Excuses. Tesla's FSD has made almost no progress over the last year. Trying to explain away why the system works the way it doesn't does not negate the fact that it has serious flaws and people should not be beta testing it with their children.
When it confuses a horse and cart for a pickup truck, semi truck, and a person walking on the road all within 30 seconds, or confuses a child for a cone, you have serious problems and it should be in no way testing your buggy software on public roads.
Question: why do they split it by 1:3? This will cause approximation of the share prices. E.g. If you bought 10 shares at $100, it would become 30 shares at $33.33333333... which is rounded to $33.33, a loss, esp. if you bought many?
Question: why do they split it by 1:3? This will cause approximation of the share prices. E.g. If you bought 10 shares at $100, it would become 30 shares at $33.33333333... which is rounded to $33.33, a loss, esp. if you bought many?
There isn’t a loss and you don’t understand how splits work.
Please register to post and access all features of our very popular forum. It is free and quick. Over $68,000 in prizes has already been given out to active posters on our forum. Additional giveaways are planned.
Detailed information about all U.S. cities, counties, and zip codes on our site: City-data.com.
Please register to participate in our discussions with 2 million other members - it's free and quick! Some forums can only be seen by registered members. After you create your account, you'll be able to customize options and access all our 15,000 new posts/day with fewer ads.
