LOL - funniest is when you drive on the highway and a sign reminds you that the MINIMUM speed is 40 mph. But FSD immediately adjust the SPEED LIMIT to 40 mph. Then, 10-20 seconds later, somehow the Google Maps metadata prevails and the speed limit is re-adjusted to 70 mph. Then.... another 40 mph MINIMUM speed appears and FSD, again, resets to speed limit to 40 mph.
@@TCPUDPATM Seems to happen every single time a minimum speed sign is posted. FSD 12.3.6. Also, sometimes car appears unable to distinguish 5 from 6 on speed limit signs. So a 65 mph limit is interpreted as 55 mph.
Dude just calm the f down unless you are part of the fsd development team or can create this all on your own. It's still years ahead of other brands out there. Just stfu n observe its evolution whether it's slow or fast
The problem is that earlier versions didn't have this major safety issue. This should be an immediate recall. This is a degradation of quality of FSD. Also, this was most frustrating because over four weeks the issue did not improve at all.
After *four* *long* *years* in beta, FSD is still primitive and very prone to errors on any challenging drive. FSD cannot be trusted, and full autonomy is years away .
@@blacknbluereign8567 Imagine your driver driving you 60 on a 50 mph road and suddenly slowing down to 15 mph, time and time again. If you don't get angry at your driver for consistently doing that, will your driver ever learn?
The only way this will ever work in every situation and scenari is when it has general intelligence equivalent to human and probably carrying with it the definition of what it means to be sentient/ conscious. If it ever got to that state you'd surely have to ask its permission to drive you round . Otherwise thats slavery . If you did, you might find it had better things to do. For me, thats the ultimate reason why FSD will always fail./
@@ElectricBrian I remain unconvinced. What about all the almost inifinite varaiations of situations that can emerge. Machine learning can learn it all - given infinite time and infinite processing power -but thats not realistic. Consider a dog running out on the street. Most people would emergency brake or attempt to sharply steer around to avoid. Now consider a plastic bag blowing out in the wind. Most people can recognise in milliseconds and just run over - without overreacting on the brake. Now consider a dog with a plastic bag attched to its collar. Then consider a blowing bag with a dog's picture. You can invent thousands of possibilities just for this one scenario. It might be unlikley but it could happen once in your lifetime as a new event. And many more more unlikely events could as well. Most humans can make quick decision and react appropriately in most novel situations . Maybe FSD can - but not seolely via machine learning approach - it has to process genuine new unecountered scenarios. I believe only by accumulating as much intelligence and as much life expeience as an average human could it do that. And that may never happen.
@@bbbf09 I agree with your assessment - there exists a vast number of unanticipated "edge cases". Like, the whole world is edge cases. The moral problem of enslaving a sentient being is interesting, but is there any reliable way to assess sentience (like a legit Turing test)?
LOL - funniest is when you drive on the highway and a sign reminds you that the MINIMUM speed is 40 mph. But FSD immediately adjust the SPEED LIMIT to 40 mph. Then, 10-20 seconds later, somehow the Google Maps metadata prevails and the speed limit is re-adjusted to 70 mph. Then.... another 40 mph MINIMUM speed appears and FSD, again, resets to speed limit to 40 mph.
Wow, that's super dangerous, too! This is why having to press the accelerator needs to be counted as a disengagement.
Wow! I’ve never had that happen.
@@TCPUDPATM Seems to happen every single time a minimum speed sign is posted. FSD 12.3.6.
Also, sometimes car appears unable to distinguish 5 from 6 on speed limit signs. So a 65 mph limit is interpreted as 55 mph.
Electric Brian is yelling at a car with an Electric Brain.
F minus Brian. F-
lol
@@idroppedmypocket What, I thought Elon was listening!
Dude just calm the f down unless you are part of the fsd development team or can create this all on your own. It's still years ahead of other brands out there. Just stfu n observe its evolution whether it's slow or fast
"Slow" is the operative word here.
The problem is that earlier versions didn't have this major safety issue. This should be an immediate recall. This is a degradation of quality of FSD. Also, this was most frustrating because over four weeks the issue did not improve at all.
After *four* *long* *years* in beta, FSD is still primitive and very prone to errors on any challenging drive.
FSD cannot be trusted, and full autonomy is years away .
decades away
@@bbbf09 Agreed. FSD is a bumbling beta software with a lack of situational awareness. A human driver is in a whole different league.
Imagine throwing a fit cause your car isn't driving you perfectly. The Car is DRIVING ITSELF!!
@@blacknbluereign8567 Imagine your driver driving you 60 on a 50 mph road and suddenly slowing down to 15 mph, time and time again. If you don't get angry at your driver for consistently doing that, will your driver ever learn?
The only way this will ever work in every situation and scenari is when it has general intelligence equivalent to human and probably carrying with it the definition of what it means to be sentient/ conscious. If it ever got to that state you'd surely have to ask its permission to drive you round . Otherwise thats slavery . If you did, you might find it had better things to do.
For me, thats the ultimate reason why FSD will always fail./
I sure hope it gets so much better in the next couple years. Some of the main issues I've had seem relatively easy to fix.
@@ElectricBrian Chevy's Supercruise is much better on the highways
@@ElectricBrian I remain unconvinced. What about all the almost inifinite varaiations of situations that can emerge. Machine learning can learn it all - given infinite time and infinite processing power -but thats not realistic. Consider a dog running out on the street. Most people would emergency brake or attempt to sharply steer around to avoid. Now consider a plastic bag blowing out in the wind. Most people can recognise in milliseconds and just run over - without overreacting on the brake. Now consider a dog with a plastic bag attched to its collar. Then consider a blowing bag with a dog's picture. You can invent thousands of possibilities just for this one scenario. It might be unlikley but it could happen once in your lifetime as a new event. And many more more unlikely events could as well. Most humans can make quick decision and react appropriately in most novel situations . Maybe FSD can - but not seolely via machine learning approach - it has to process genuine new unecountered scenarios. I believe only by accumulating as much intelligence and as much life expeience as an average human could it do that. And that may never happen.
@@bbbf09 I agree with your assessment - there exists a vast number of unanticipated "edge cases". Like, the whole world is edge cases. The moral problem of enslaving a sentient being is interesting, but is there any reliable way to assess sentience (like a legit Turing test)?
You need to calm down.
@@SteveHischar You're absolutely right. Now that I don't have this bad version of FSD, I am much calmer.