3 ms·

Yes he has over-promised, but Tesla also has the most promising tech on this for many reasons. It seems more like the problem itself is turning out to just be h

by maxlin 3y ago

Yes he has over-promised, but Tesla also has the most promising tech on this for many reasons. It seems more like the problem itself is turning out to just be harder to cover properly than foreseen originally. However, cars last long and Tesla's batteries last as long as any, I remember seeing some very promising numbers from very old Model S's remaining battery capabilities, and of course the batteries can be replaced. I don't see why the robotaxi idea wouldn't increase the potential value of almost any Tesla considering even the NN chip hardware can be semi-easily upgraded if really needed > that can never exist. That sounds like someone saying muskets are a fad. Do you really truly believe that? Never heard anyone be _that_ pessimistic about it, especially with the continual improvements.

IshKebab 3y ago
> the problem itself is turning out to just be harder to cover properly than foreseen originally Than foreseen by Tesla. Everyone else knew they couldn't possibly achieve what Musk was claiming. > Tesla also has the most promising tech on this for many reasons After Waymo maybe. > considering even the NN chip hardware can be semi-easily upgraded if really needed Haha there's no way 20 year old cars (minimum; by the time FSD actually works) are going to get sensor and compute upgrades. Not going to happen. > That sounds like someone saying muskets are a fad. I'm not sure what your analogy is meant to mean but I think you misunderstood me. I'm not saying driverless cars in general can never exist. Clearly they will eventually (maybe 10-20 years?). It's the upgrade to FSD that people have paid for that can never exist.
- maxlin 3y ago
  You are dumbfoundingly incorrect. Tesla is just so much ahead everyone else it is a bit ridiculous. I wouldn't wonder if many have already given up and are just waiting for Tesla to front the R&D then willing to license their tech. No one else has even close to Teslas NN chips, Dojo, custom simulators, the FLEET, willing free QA people, endless data sources as all cars have the hardware ... You also have no idea how time works throwing around something like 20 years. In AI age that just leads me to believe you are at least 10 years late in your thinking. 20 year guess for software is something no one can make. You do not even understand what field Tesla operates in if you're comparing it to Waymo. Waymo operates on HD maps and remote human operators who can take over every 5 minutes if needed. Those cars are not real "cars", you can't own them, they can't drive outside the city. They compete with electric scooters like Tier, Bolt, Lime, not Tesla. They have zero ability to grow, it's like comparing a hardcoded nature language engine to ChatGPT. (Also note that the gain in value Tesla has gotten after Autonomy Day hasn't actually even been lost. Long-term investors aren't gonna miss this one)
martin8412 3y ago
Tesla is nowhere near the leader. The problem is hard, who could possibly have known? Oh right, any person with a bit of knowledge of how computers work and drive a car. Tesla can't make it reliably work in the car-centric USA. I'd love to see it perform in a busy Indian city or even just any old European city. Tesla cars do not last long, and the battery costs a fortune to replace when it dies or gets damaged out of warranty.
- maxlin 3y ago
  Lol what. It's like you'd be trying to convince everyone Bing has been the leading search engine for the last 10 years. You don't even mention who would be ahead. Cruise, having to intervene every 5 minutes (literally, look it up) and basing on unscalable HD maps? Get real man. Tesla has even custom HARDWARE for their NN's even, designed by the chip designer that has created some of the greatest designs BOTH at Intel and AMD. And Tesla's battery lasts better than others. I remember seeing incredible endurance even for the first Model S batteries after like 10 years of driving. You buy a Tesla now or 5 years ago, it will likely last longer than any other car, all while it will receive SW updates to function as a Robotaxi when the time comes. You're funnily arguing the exact polar opposite of where Tesla is strong.
  - IshKebab 3y ago
    > Tesla has even custom HARDWARE for their NN's Their chips are actually nice, but you're betraying your lack of knowledge here - the compute hardware isn't remotely the hard part about self driving cars.
    - maxlin 3y ago
      Using words like "remotely not the hard part" that is just not true man. There doesn't exist hardware that can do inference with even close to as low as the 100 watts Tesla uses. Also it is not just to drive FSD, it is to provide data, to detect situations of from where to send data as they can't just stream everything to base (check the Autonomy day talk, it's quite interesting), which is needed for the development of the software which we agree is an extremely hard problem. To further demonstrate this, consider the fact that in the current AI landscape, hardware is actually the bottleneck. Everyone is trying to buy more GPU's NVIDIA simply can't supply fast enough. If Tesla still relied on NVIDIA for the in-car processing, the project would have to be shelved as they'd have to sell Teslas without the hardware, instead of just having it all cars even if they will never use it for running FSD directly on that car.

by maxlin 3y ago

IshKebab 3y ago

> the problem itself is turning out to just be harder to cover properly than foreseen originally Than foreseen by Tesla. Everyone else knew they couldn't possibly achieve what Musk was claiming. > Tesla also has the most promising tech on this for many reasons After Waymo maybe. > considering even the NN chip hardware can be semi-easily upgraded if really needed Haha there's no way 20 year old cars (minimum; by the time FSD actually works) are going to get sensor and compute upgrades. Not going to happen. > That sounds like someone saying muskets are a fad. I'm not sure what your analogy is meant to mean but I think you misunderstood me. I'm not saying driverless cars in general can never exist. Clearly they will eventually (maybe 10-20 years?). It's the upgrade to FSD that people have paid for that can never exist.

maxlin 3y ago
You are dumbfoundingly incorrect. Tesla is just so much ahead everyone else it is a bit ridiculous. I wouldn't wonder if many have already given up and are just waiting for Tesla to front the R&D then willing to license their tech. No one else has even close to Teslas NN chips, Dojo, custom simulators, the FLEET, willing free QA people, endless data sources as all cars have the hardware ... You also have no idea how time works throwing around something like 20 years. In AI age that just leads me to believe you are at least 10 years late in your thinking. 20 year guess for software is something no one can make. You do not even understand what field Tesla operates in if you're comparing it to Waymo. Waymo operates on HD maps and remote human operators who can take over every 5 minutes if needed. Those cars are not real "cars", you can't own them, they can't drive outside the city. They compete with electric scooters like Tier, Bolt, Lime, not Tesla. They have zero ability to grow, it's like comparing a hardcoded nature language engine to ChatGPT. (Also note that the gain in value Tesla has gotten after Autonomy Day hasn't actually even been lost. Long-term investors aren't gonna miss this one)

martin8412 3y ago

Tesla is nowhere near the leader. The problem is hard, who could possibly have known? Oh right, any person with a bit of knowledge of how computers work and drive a car. Tesla can't make it reliably work in the car-centric USA. I'd love to see it perform in a busy Indian city or even just any old European city. Tesla cars do not last long, and the battery costs a fortune to replace when it dies or gets damaged out of warranty.

maxlin 3y ago
Lol what. It's like you'd be trying to convince everyone Bing has been the leading search engine for the last 10 years. You don't even mention who would be ahead. Cruise, having to intervene every 5 minutes (literally, look it up) and basing on unscalable HD maps? Get real man. Tesla has even custom HARDWARE for their NN's even, designed by the chip designer that has created some of the greatest designs BOTH at Intel and AMD. And Tesla's battery lasts better than others. I remember seeing incredible endurance even for the first Model S batteries after like 10 years of driving. You buy a Tesla now or 5 years ago, it will likely last longer than any other car, all while it will receive SW updates to function as a Robotaxi when the time comes. You're funnily arguing the exact polar opposite of where Tesla is strong.
- IshKebab 3y ago
  > Tesla has even custom HARDWARE for their NN's Their chips are actually nice, but you're betraying your lack of knowledge here - the compute hardware isn't remotely the hard part about self driving cars.
  - maxlin 3y ago
    Using words like "remotely not the hard part" that is just not true man. There doesn't exist hardware that can do inference with even close to as low as the 100 watts Tesla uses. Also it is not just to drive FSD, it is to provide data, to detect situations of from where to send data as they can't just stream everything to base (check the Autonomy day talk, it's quite interesting), which is needed for the development of the software which we agree is an extremely hard problem. To further demonstrate this, consider the fact that in the current AI landscape, hardware is actually the bottleneck. Everyone is trying to buy more GPU's NVIDIA simply can't supply fast enough. If Tesla still relied on NVIDIA for the in-car processing, the project would have to be shelved as they'd have to sell Teslas without the hardware, instead of just having it all cars even if they will never use it for running FSD directly on that car.