-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
When will privately owned Tesla Robotaxis be operating on HW3 or lower?
- Thread starter hooty
- Start date
SalisburySam
Active Member
what are the 2025 people smoking? 0% until 2025. Can it be called a lie yet, or is the scam still aspirational?
Is it to much to ask for specs and a roadmap for delivery on promissed car parts purchased over 4 years ago?
What is FSD and when will I get it?
Is it to much to ask for specs and a roadmap for delivery on promissed car parts purchased over 4 years ago?
What is FSD and when will I get it?
I voted "never", but that was because the poll is predicated on the "robotaxi" idea. I don't see safety agencies and insurance companies allowing the robotaxi model on something like HW3. They're going to want such huge margins of safety before turning this into a business that even HW4 may not suffice. There's just no reason for them to allow the robotaxi concept to advance quickly. Autonomy alone suffices to save lives and keep the accident rate down.
That said, I can easily see people being driven around in their own cars without intervention within a couple more years. Just some time on 11.4.2 should demonstrate that. HW3 doesn't have perception problems. It has control problems. That's on Tesla's engineering team. The slow pace of development is because this is a research project and they haven't yet found the special sauce that's needed. I don't believe that anyone has any idea what HW3 is actually capable of because the field is so new. Right now, Tesla is just trying to make it work. While I'm sure they're also trying to optimize their computes as best they can, they have yet to begin optimizing the solution itself.
johnmodely
Member
It does have perception problems with the current camera setup. I dont think HW3 cameras will ever be robotaxi capable. At minimum front bumper cameras are needed or much wider higher resolution front camera. This whole "creeping" business is super unsafe and is easily solvable with bumper cameras.
DanCar
Active Member
Ray Kurzweil predicted AGI by 2029. So I voted 2030.
AGI - artificial general intelligence
thenextweb.com
Quote:
Ray Kurzweil, a fabled futurist, loves making predictions — and they’re admirably precise. At the 2017 SXSW Conference in Austin, Texas, Kurzweil gave a typically pinpoint prediction.
“By 2029, computers will have human-level intelligence,” he said. “That leads to computers having human intelligence, our putting them inside our brains, connecting them to the cloud, expanding who we are. Today, that’s not just a future scenario. It’s here, in part, and it’s going to accelerate.”
AGI - artificial general intelligence
When will AGI arrive? Here’s what our tech lords predict
DeepMind's CEO has revealed new predictions for the arrival of AGI. Here's how they compare to the forecasts of other tech titans.
thenextweb.com
Ray Kurzweil, a fabled futurist, loves making predictions — and they’re admirably precise. At the 2017 SXSW Conference in Austin, Texas, Kurzweil gave a typically pinpoint prediction.
“By 2029, computers will have human-level intelligence,” he said. “That leads to computers having human intelligence, our putting them inside our brains, connecting them to the cloud, expanding who we are. Today, that’s not just a future scenario. It’s here, in part, and it’s going to accelerate.”
My apologies for being unclear. Certainly it has perception problems. My point was that the primary impediment to autonomy with Teslas is the control software. It has sufficient perception to drive autonomously. Not as well as HW4 will because that generation of hardware was designed based on lessons that Tesla learned from HW3, but it will be able to get around.
Would I get a HW4 upgrade if it was available for my Tesla? Yes.
Its clear from the poll results that Musk is not trustworthy
Maybe I should poll that question, too
Maybe I should poll that question, too
My god, it’s almost like people learn new things over time and change their opinions. Cutting edge research.
1) Everyone in 2017 thought self driving was going to be easy. This is not hard to validate for yourself with Google. They were all wrong including Musk.
2) Being overoptimistic is not lying if the belief is sincere.
3) Musk is very all-or-nothing. Statements that should have been wrapped in ‘hope’ and ‘aim to’ disclaimers were instead made as commitments. This is dumb but that’s how he rolls.
1) Everyone in 2017 thought self driving was going to be easy. This is not hard to validate for yourself with Google. They were all wrong including Musk.
2) Being overoptimistic is not lying if the belief is sincere.
3) Musk is very all-or-nothing. Statements that should have been wrapped in ‘hope’ and ‘aim to’ disclaimers were instead made as commitments. This is dumb but that’s how he rolls.
Disagree. Despite all the missteps and bad prognostications, I believe Tesla will obtain L3 autonomy on HW3. Really, the only thing standing in the way of this right now is Elon’s L5 pipe dreams.
SalisburySam
Active Member
I remain skeptical, but I appreciate your optimism.
We fully agree here, and the reason for my earlier skepticism. Musk seems unmovingly attached to that dream, reality (cost?) be damned. My Model 3 is a terrific car, and I can only imagine what it could have been today had the software development not been cluttered with its robotaxi silliness.
Last edited:
Times have changed meaning what? In this subforum here? This subforum has been extremely pessimistic about FSD pretty much since I joined here in 2018.
No, it really hasnt, not since I joined here in 2018. This subforum is pretty much entirely comprised of people who talk about how poor FSD is, with just a few people who push back on that. Almost anyone who actually thought positive about FSD stopped discussing it here a long time ago, with a few holdouts.
SalisburySam
Active Member
I didnt say whether it was justified or not, I have no opinion on that whatsoever. The only thing I was expressing an opinion on was the notion that "wow where did all the pessimism come from, thats new!" when it absolutely isnt, at least not since 2018.
SalisburySam
Active Member
Understand. I think more than a few of us who opted in on FSD half a decade ago and have seen price increases but little to show for it are feeling quite negative today. Certainly that’s how I feel and the source of at least my pessimism. It has been building up given the tenor of this and other Tesla forums, so maybe not all that new. But yes, this is a big reversal of the enthusiasm of earlier days. Lastly, the intractability of solving seemingly simple things like phantom braking and questionable wiper controls among others leads at least one of us to question Tesla’s ability to solve the much more difficult issues of L3 and beyond, or perhaps their priority and will to do so.
I know I’ve become pretty negative on this forum though I do enjoy most of my Model 3 experiences. And there indeed several areas in which Tesla has succeeded in truly amazing ways including the charging network, the incredible heat pump system that scavenges heat from 16 sources throughout the vehicle, and the quiet and convenience of a very dependable EV. However, my car is intentionally stuck in time to mid-2022 running on v2022.20.8, the last update I installed prior to the one that murdered my paid-for and working radar in addition to other downgrades. Yeah, can generate some pessimism.
Last edited:
How are Tesla supposed to give you ‘something to show for’ the money you’ve paid if you refuse their updates? This seems like an odd position to adopt.
What’s given you such confidence that the old software with radar works better than the current vision software? I never owned a Tesla with radar but I had a VW with radar cruise control and the TV autopilot is undoubtably better than that. Of all the threads I’ve seen on here from radar Tesla owners who’ve updated the significant majority seem to say ‘it was ropey at first but now it’s better than radar’.
Similar threads
- Article
- Poll
