26 ms·

> a fleet of self-driving vehicles that will make their way into the robotaxi service as soon as next year. So... are we living on the same planet? Does Musk s

> a fleet of self-driving vehicles that will make their way into the robotaxi service as soon as next year. So... are we living on the same planet? Does Musk seriously think that Tesla will have solved full autonomous driving, with no LIDAR, and have commercialized a functioning robotaxi service, by next year? This is so far outside of the realm of possibility that it's like saying he's going to Pluto next week.

JohnJamesRambo 7y ago
It's either a lie or psychotic delusion, I refuse to believe he is that uninformed and ignorant about the current state of autonomous driving.
- tachyonbeam 7y ago
  I think he's extremely optimistic and confident. He's been extremely successful so far, he's a billionaire, has done many things that nay sayers would have called impossible (ie: paypal, landing rockets, building electric cars with very competitive performance). It's not hard to see why Musk might believe he can do anything. He's taken big risks and gotten a lot of positive reinforcement over the last two decades. I agree that his timeline for robotaxis is probably over optimistic, and that there are many unsolved problems. Self-driving isn't all they have to solve. They also have to solve complex and annoying problems such as having self-driving cars find the people they are supposed to pick up, and also having the cars somehow find charging, find parking, etc. Furthermore, Tesla's claim of 1 million robotaxis assumes that everyone who owns a Tesla with full self driving would want to rent them out as part of the Tesla Network. That's very unrealistic. Most people probably prefer having their car be always available and not subjecting it to potential damage/towing/fines (more driving, more risk). Still, I wouldn't say Musk is ignorant about the state of self driving. They have self driving cars which they have demoed to people on their investor day. He just makes this probably incorrect assumption that it's all "just software" from here, and everything should be easy to fix. Software isn't easier than hardware. It could also be the case that changes do need to be made for the hardware in order to get high enough reliability, there is no guarantee that the hardware is completely right in its current iteration.
  - camjohnson26 7y ago
    It’s not over optimistic, it’s delusional and will not happen. Tesla doesn’t have self driving cars, what they demoed was a less restricted version of autopilot which had to disengage multiple times on the short, predefined demo route that investors rode on. There’s an easy explanation. Musk knows that self driving is mostly hype at this point but he’s running out of cash and desperately needs a capital raise. If he can fool investors into thinking he’ll have self driving solved within a year maybe they’ll give Tesla the money it needs.
    - Someone 7y ago
      “Fooling investors” can also lead to law suits that cost Tesla and Musk lots of time and attention, and, in extreme cases, may lead to Musk spending time in jail.
- Tiktaalik 7y ago
  With regard to the Boring Company and his transit 'solutions' he's obviously uninformed and ignorant, and his bs has been called out so many times by transit planners, so I wouldn't be surprised if he's just as uninformed and ignorant with his other endeavors.
- tim333 7y ago
  I think he's counting on 'exponential improvement'. You can sort of see where they are going - they have features like summon in a car park or changing lane in traffic that sort of work and they are gradually going to tick off the various driving test skills you need at which point it will sort of work. Then he's hoping it will get reliable as the network learns. It's that last point that may be over optimistic.
  - fmihaila 7y ago
    > I think he's counting on 'exponential improvement'. Here he is, on record, stating that he believes Tesla has essentially solved the problem ('game, set, and match'): [1] https://www.youtube.com/watch?v=dEv99vxKjVI&t=1778 https://www.youtube.com/watch?v=dEv99vxKjVI&t=1778
  - manicdee 7y ago
    The car also stops at signals and stop signs, and yields for traffic before turning. Did you watch the autonomy day presentations and demonstration drives?
    - tim333 7y ago
      Yeah I did. It seems to be coming along.
overcast 7y ago
Oh come on, he's got a whole seven months!
- bryanlarsen 7y ago
  19 months from now is also next year. And "as soon as" means "maybe". And as all parents know, "maybe" means "no".
  - ericb 7y ago
    Even speaking as a fan, this is an excellent summary.
  - dwighttk 7y ago
    For what it is worth Musk said: “I think we will be feature complete, full self-driving, this year – meaning the car will be able to find you in a parking lot, pick you up and take you all the way to your destination without intervention, this year. I would say I am of certain of that. That is not a question mark.” That’s 7 months, not 19. (The rest is good :)
    - cjhopman 7y ago
      He also said "We'll be able to do a demonstration guide of full autonomy all the way from LA to New York. So basically from home in LA to Times Square in New York. And then have the car go and park itself by the end of next year"
      - dwighttk 7y ago
        when was that? 2016? :D
BrandonMarc 7y ago
He's used to overpromising and then delivering far later than expected - it's worked so well for SpaceX. Maybe he inherited Jobs' reality distortion field?
- dna_polymerase 7y ago
  > it's worked so well for SpaceX In Elon's biography the author mentions that in all projects he was part of he would constantly give far optimistic delivery dates.
  - mpweiher 7y ago
    A true engineer! :)
    - ergothus 7y ago
      I realize you're joking, but this is actually kind of bothersome to me. Most of us give overly optimistic delivery dates without intending to. Over time we learn that we have this habit, so we pad in some time we think is unnecessary...and still give overly optimistic delivery dates without intending to. If you add someone INTENDING to give such poor estimates, the problem moves from the "expected and occasionally problematic but still joke-able" to the "just plain lies" category.
      - Robotbeat 7y ago
        He's not intending to be over-optimistic; his goals are possible, he sees they are possible and says so, even if they're unlikely. And he repeatedly states his tendency to be over-optimistic, which RARELY is acknowledged by pretty much anyone else promoting a new advancement. "I think we can do this by XX. Granted, I've been over-optimistic in the past"... is refreshing honesty, not "lies." It's amazing to me that so many people interpret acknowledgement of one's own fallibility as lying.
        ergothus 7y ago
danso 7y ago
Also, have they given any evidence that they’ve thought through and prepared for the regulatory hurdles that a robotaxi service would entail, even if the tech problems were solved?
- ucaetano 7y ago
  No, on purpose. Musk is reserving "regulatory" as the excuse that will be used when they fail to deliver: "we had everything ready, I swear, but regulators wouldn't let us launch".
- CoolGuySteve 7y ago
  Why bother with that when you can go shopping for a jurisdiction that will let you do whatever you want with minimal interference? Seemed to work for Uber in Pheonix... until somebody died.
LeoPanthera 7y ago
My theory is that it's going to work in the style of Zipcar, rather than Uber. The cars won't move without anyone in them, but you'll be able to reserve one, get in the drivers seat, and then it will drive you to your destination, presumably with human assistance should the autopilot disengage.
- notfromhere 7y ago
  It's not real, it's not going to happen, and setting up even a human rideshare service is going to take longer than 12 months
  - toomuchtodo 7y ago
    I have already delegated access to my Model S to other Tesla account holders through customer service (manual process). Seems to automate it is straightforward? Tesla could replace Turo for Teslas fairly rapidly IMHO (and possibly Zipcar!). Supporting Lyft/Uber functionality would of course be more effort, but also doable. This isn't trying to go to the Mars (insert chuckle here). Even if you discount full autonomy, Tesla could collect a skim from every one of these transactions, while also collecting autopilot data from each trip. Here's an open source rideshare platform: https://libretaxi.org/ https://libretaxi.org/ TLDR Lots of dollars to capture besides for handing manufactured units over to customers. It's so silly people value Uber and Lyft as platforms so highly (as just mobile apps) while they have so little capability compared to Tesla, who not only has similar software engineering capabilities, but actually builds the cars and restricts their ability to join other platforms as deeply integrated components.
    - notfromhere 7y ago
      So Tesla is going to enter an industry that actively loses money based on an autonomy play that doesn't exist and isn't even legal on most U.S. roads?
      - hef19898 7y ago
        Exactly what you would do to solve your cash flow and profit issues, isn't it?
CoolGuySteve 7y ago
A robotaxi isn't too far off if you limit where it can operate. It's not infeasible to train an automated driving system specifically suited to the streets of LA or wherever. Some place where it doesn't snow, doesn't rain often, and drivers/cyclists/pedestrians are fairly predictable (at least compared to someplace like Asia). Alternatively, a robobus, that is to say an automated system that drives from one place to the other on a highway or interstate, seems to be mostly solved. In both cases, if the vehicles were mostly autonomous but could also be controlled remotely by a human in ambiguous cases, they would still be vastly cheaper to operate than existing taxi/bus services.
- dragontamer 7y ago
  Is Tesla really ready for a long protracted war against Uber, Waymo (aka Google), and other autonomous taxi companies? Tesla just spent (probably) ~$100 Million or so developing a 14nm chip, already obsolete (NVidia has a 12nm chip already, Intel/MobilEye has 7nm incoming 2020, and who knows what Google/Waymo got up their sleeves. Those TPUs are fully proprietary and fully to Google/Waymo's advantage). 7nm is deployed and shipping, Tesla's already behind. And now you're telling me that to "raise money", they're entering another highly competitive field full of companies who are literally willing to lose money over the next 5-years on this idea? That's straight up madness. I'm sure autonomous taxis will eventually become a reality, but no company is going to make money in that field for 5 to 10 years. The other companies like Google / Waymo are going to do what Silicon Valley does best: loss lead for years and years to build a customer base, before raising prices and eventually making a sustainable business. Tesla has already spent all of its cash and loans on Gigafactory 1, Gigafactory 2, and Gigafactory 3. Scratch that, Gigafactory 3 isn't even complete yet (and is estimated to be a few billion dollars alone in costs...). At best, Tesla will have to enter the autonomous taxi industry offering goods at cost (or higher: trying to make a profit). That's innately going to put them severely behind Google / Waymo, who will loss lead their way to cheaper prices and more customers.
  - DennisP 7y ago
    There's more to it than feature size. Other companies have more general-purpose chips. Tesla basically built an ASIC for self-driving.
Nokinside 7y ago
It seems to be completely pie in the sky type of thinking. The only rational -non delusional- reason I see for the talk is if Musk does it to get better financing terms before the self driving hype dries down. I think taxi services without driver are coming but the the secret sauce is remote human operator working together with high level of automation. Driver is the main cost of taxi service. If you get one operator to remotely oversee 20 highly autonomous vehicles and provide input when needed, the cost of taxi-service drops into a fraction of what it used to be and service is better than fully autonomous driving.
- izzydata 7y ago
  I would never trust such a ridiculous system. I don't know that many other people would either. Although I would also not trust a fully autonomous vehicle either until more evidence is provided that we were even within 20 years of it.
- notfromhere 7y ago
  So humans that can barely drive a single vehicle will now be responsible for monitoring 20 vehicles in highly different environments that change at a moments notice?
  - Nokinside 7y ago
    Not at a moment's notice. The automation must react during emergencies, but humans can navigate and solve stuff that the automation can't.
    - marsokod 7y ago
      Yes. The car is responsible for the safety in all circumstances and recognizes it's limitations. Think level 4 autonomy. As soon as the car sees it cannot proceed properly, it parks somewhere or go in a safe situation and requests a remote human operator to intervene. Then the whole point is to have a call center sized so that you do not have to wait too much. We're not there yet, and giving back the control to the passengers is also a possibility. However, in a world where passengers almost never drive, it might be safer to give the control to trained remote operators.
ajmarsh 7y ago
I don't think it's that far off at all. https://x-matik.com/ https://x-matik.com/
mruts 7y ago
Is it fair to ask at this point if Elon Musk is going crazy? His decisions of late seem unwise at best, irrational at worst. Maybe he should just concentrate on producing Model 3's and then the Model Y? Or what would be really cool would if Tesla could produce multiple car models... at the same time!
- elorant 7y ago
  More like desperate. He has to hit targets that seem impossible otherwise he'll run out of cash.
DennisP 7y ago
I've always been skeptical of skipping LIDAR, but a recent study at Cornell found that with binocular vision they could be almost as accurate as LIDAR, but with much cheaper hardware. https://www.therobotreport.com/researchers-back-teslas-non-lidar-approach-to-self-driving-cars/ https://www.therobotreport.com/researchers-back-teslas-non-l...
- Judgmentality 7y ago
  That study has been highly disputed by people in the field; the methodology is questionable at best for the claims they're making.
  - toomuchtodo 7y ago
    Anthony Levandowski of Waymo LIDAR fame has said he now agrees with Elon Musk ("a wiser person than me") that LIDAR is a crutch. https://youtu.be/fNgEG5rCav4?t=105 https://youtu.be/fNgEG5rCav4?t=105
    - xmprt 7y ago
      I wonder why the person who was taken to court over (allegedly) stealing Waymo's IP related to LIDAR would be wary of using LIDAR in his new projects...
    - Judgmentality 7y ago
      Anthony Levandowski is so sleazy he makes Vinod Khosla look like a good guy. He is (well, was) one of the most untrustworthy people in the Valley, so much so he had to leave for China to start his next company because even local VCs - who care about money and nothing but money - found him too sketchy to work with.
      - toomuchtodo 7y ago
        You can be sleazy and untrustworthy and still be smart and right. I wouldn't use VC funding as signaling, considering how many VC folks blindly funded Theranos (which wasn't just outright fraud, but abusive pursuit of whistleblowers). It's fairly clear machine vision is going to surpass LIDAR in the near future (taking capability and cost into account), doesn't take a robotics expert to see that.
agumonkey 7y ago
You don't buy his claims that lidar is non necessary and even inefficient at sdv ?
- mplewis 7y ago
  Elon/Tesla have a vested interest in making sure LIDAR isn't required by law for self-driving vehicles. In addition, his NN-camera-powered-cars are still trying to drive themselves into highway barriers.
  - Robotbeat 7y ago
    Requiring a specific technical solution with problems of its own by law is a wet dream for whoever holds the IP to that technical solution. Now accusing others of having a vested interest in MAYBE opposing this hypothetical law that no one has actually proposed is just gas-lighting and possibly projection.
    - notfromhere 7y ago
      LiDAR means he'd have to retrofit vehicles, given how he promised every purchaser that their cars are self-driving ready. LiDAR literally will cost him money, so yeah, that's a conflict.
  - agumonkey 7y ago
    Yeah but back to its point, he argued lidar was not optimal for "computer vision", do you think it has limits that would be shadowed by multi camera NN analysis ? That he's doing bad PR is another point (that we can discuss also but is harder to comment on objectively)
  - manicdee 7y ago
    Do you believe lidar is so necessary that it must be mandated by law?
senectus1 7y ago
He's not the only one working on this. I think its seriously deliverable.

JohnJamesRambo 7y ago

It's either a lie or psychotic delusion, I refuse to believe he is that uninformed and ignorant about the current state of autonomous driving.

tachyonbeam 7y ago
I think he's extremely optimistic and confident. He's been extremely successful so far, he's a billionaire, has done many things that nay sayers would have called impossible (ie: paypal, landing rockets, building electric cars with very competitive performance). It's not hard to see why Musk might believe he can do anything. He's taken big risks and gotten a lot of positive reinforcement over the last two decades. I agree that his timeline for robotaxis is probably over optimistic, and that there are many unsolved problems. Self-driving isn't all they have to solve. They also have to solve complex and annoying problems such as having self-driving cars find the people they are supposed to pick up, and also having the cars somehow find charging, find parking, etc. Furthermore, Tesla's claim of 1 million robotaxis assumes that everyone who owns a Tesla with full self driving would want to rent them out as part of the Tesla Network. That's very unrealistic. Most people probably prefer having their car be always available and not subjecting it to potential damage/towing/fines (more driving, more risk). Still, I wouldn't say Musk is ignorant about the state of self driving. They have self driving cars which they have demoed to people on their investor day. He just makes this probably incorrect assumption that it's all "just software" from here, and everything should be easy to fix. Software isn't easier than hardware. It could also be the case that changes do need to be made for the hardware in order to get high enough reliability, there is no guarantee that the hardware is completely right in its current iteration.
- camjohnson26 7y ago
  It’s not over optimistic, it’s delusional and will not happen. Tesla doesn’t have self driving cars, what they demoed was a less restricted version of autopilot which had to disengage multiple times on the short, predefined demo route that investors rode on. There’s an easy explanation. Musk knows that self driving is mostly hype at this point but he’s running out of cash and desperately needs a capital raise. If he can fool investors into thinking he’ll have self driving solved within a year maybe they’ll give Tesla the money it needs.
  - Someone 7y ago
    “Fooling investors” can also lead to law suits that cost Tesla and Musk lots of time and attention, and, in extreme cases, may lead to Musk spending time in jail.
Tiktaalik 7y ago
With regard to the Boring Company and his transit 'solutions' he's obviously uninformed and ignorant, and his bs has been called out so many times by transit planners, so I wouldn't be surprised if he's just as uninformed and ignorant with his other endeavors.
tim333 7y ago
I think he's counting on 'exponential improvement'. You can sort of see where they are going - they have features like summon in a car park or changing lane in traffic that sort of work and they are gradually going to tick off the various driving test skills you need at which point it will sort of work. Then he's hoping it will get reliable as the network learns. It's that last point that may be over optimistic.
- fmihaila 7y ago
  > I think he's counting on 'exponential improvement'. Here he is, on record, stating that he believes Tesla has essentially solved the problem ('game, set, and match'): [1] https://www.youtube.com/watch?v=dEv99vxKjVI&t=1778 https://www.youtube.com/watch?v=dEv99vxKjVI&t=1778
- manicdee 7y ago
  The car also stops at signals and stop signs, and yields for traffic before turning. Did you watch the autonomy day presentations and demonstration drives?
  - tim333 7y ago
    Yeah I did. It seems to be coming along.

overcast 7y ago

Oh come on, he's got a whole seven months!

bryanlarsen 7y ago
19 months from now is also next year. And "as soon as" means "maybe". And as all parents know, "maybe" means "no".
- ericb 7y ago
  Even speaking as a fan, this is an excellent summary.
- dwighttk 7y ago
  For what it is worth Musk said: “I think we will be feature complete, full self-driving, this year – meaning the car will be able to find you in a parking lot, pick you up and take you all the way to your destination without intervention, this year. I would say I am of certain of that. That is not a question mark.” That’s 7 months, not 19. (The rest is good :)
  - cjhopman 7y ago
    He also said "We'll be able to do a demonstration guide of full autonomy all the way from LA to New York. So basically from home in LA to Times Square in New York. And then have the car go and park itself by the end of next year"
    - dwighttk 7y ago
      when was that? 2016? :D

BrandonMarc 7y ago

He's used to overpromising and then delivering far later than expected - it's worked so well for SpaceX. Maybe he inherited Jobs' reality distortion field?

dna_polymerase 7y ago
> it's worked so well for SpaceX In Elon's biography the author mentions that in all projects he was part of he would constantly give far optimistic delivery dates.
- mpweiher 7y ago
  A true engineer! :)
  - ergothus 7y ago
    I realize you're joking, but this is actually kind of bothersome to me. Most of us give overly optimistic delivery dates without intending to. Over time we learn that we have this habit, so we pad in some time we think is unnecessary...and still give overly optimistic delivery dates without intending to. If you add someone INTENDING to give such poor estimates, the problem moves from the "expected and occasionally problematic but still joke-able" to the "just plain lies" category.
    - Robotbeat 7y ago
      He's not intending to be over-optimistic; his goals are possible, he sees they are possible and says so, even if they're unlikely. And he repeatedly states his tendency to be over-optimistic, which RARELY is acknowledged by pretty much anyone else promoting a new advancement. "I think we can do this by XX. Granted, I've been over-optimistic in the past"... is refreshing honesty, not "lies." It's amazing to me that so many people interpret acknowledgement of one's own fallibility as lying.
      - ergothus 7y ago

danso 7y ago

Also, have they given any evidence that they’ve thought through and prepared for the regulatory hurdles that a robotaxi service would entail, even if the tech problems were solved?

ucaetano 7y ago
No, on purpose. Musk is reserving "regulatory" as the excuse that will be used when they fail to deliver: "we had everything ready, I swear, but regulators wouldn't let us launch".
CoolGuySteve 7y ago
Why bother with that when you can go shopping for a jurisdiction that will let you do whatever you want with minimal interference? Seemed to work for Uber in Pheonix... until somebody died.

LeoPanthera 7y ago

My theory is that it's going to work in the style of Zipcar, rather than Uber. The cars won't move without anyone in them, but you'll be able to reserve one, get in the drivers seat, and then it will drive you to your destination, presumably with human assistance should the autopilot disengage.

notfromhere 7y ago
It's not real, it's not going to happen, and setting up even a human rideshare service is going to take longer than 12 months
- toomuchtodo 7y ago
  I have already delegated access to my Model S to other Tesla account holders through customer service (manual process). Seems to automate it is straightforward? Tesla could replace Turo for Teslas fairly rapidly IMHO (and possibly Zipcar!). Supporting Lyft/Uber functionality would of course be more effort, but also doable. This isn't trying to go to the Mars (insert chuckle here). Even if you discount full autonomy, Tesla could collect a skim from every one of these transactions, while also collecting autopilot data from each trip. Here's an open source rideshare platform: https://libretaxi.org/ https://libretaxi.org/ TLDR Lots of dollars to capture besides for handing manufactured units over to customers. It's so silly people value Uber and Lyft as platforms so highly (as just mobile apps) while they have so little capability compared to Tesla, who not only has similar software engineering capabilities, but actually builds the cars and restricts their ability to join other platforms as deeply integrated components.
  - notfromhere 7y ago
    So Tesla is going to enter an industry that actively loses money based on an autonomy play that doesn't exist and isn't even legal on most U.S. roads?
    - hef19898 7y ago
      Exactly what you would do to solve your cash flow and profit issues, isn't it?

CoolGuySteve 7y ago

A robotaxi isn't too far off if you limit where it can operate. It's not infeasible to train an automated driving system specifically suited to the streets of LA or wherever. Some place where it doesn't snow, doesn't rain often, and drivers/cyclists/pedestrians are fairly predictable (at least compared to someplace like Asia). Alternatively, a robobus, that is to say an automated system that drives from one place to the other on a highway or interstate, seems to be mostly solved. In both cases, if the vehicles were mostly autonomous but could also be controlled remotely by a human in ambiguous cases, they would still be vastly cheaper to operate than existing taxi/bus services.

dragontamer 7y ago
Is Tesla really ready for a long protracted war against Uber, Waymo (aka Google), and other autonomous taxi companies? Tesla just spent (probably) ~$100 Million or so developing a 14nm chip, already obsolete (NVidia has a 12nm chip already, Intel/MobilEye has 7nm incoming 2020, and who knows what Google/Waymo got up their sleeves. Those TPUs are fully proprietary and fully to Google/Waymo's advantage). 7nm is deployed and shipping, Tesla's already behind. And now you're telling me that to "raise money", they're entering another highly competitive field full of companies who are literally willing to lose money over the next 5-years on this idea? That's straight up madness. I'm sure autonomous taxis will eventually become a reality, but no company is going to make money in that field for 5 to 10 years. The other companies like Google / Waymo are going to do what Silicon Valley does best: loss lead for years and years to build a customer base, before raising prices and eventually making a sustainable business. Tesla has already spent all of its cash and loans on Gigafactory 1, Gigafactory 2, and Gigafactory 3. Scratch that, Gigafactory 3 isn't even complete yet (and is estimated to be a few billion dollars alone in costs...). At best, Tesla will have to enter the autonomous taxi industry offering goods at cost (or higher: trying to make a profit). That's innately going to put them severely behind Google / Waymo, who will loss lead their way to cheaper prices and more customers.
- DennisP 7y ago
  There's more to it than feature size. Other companies have more general-purpose chips. Tesla basically built an ASIC for self-driving.

Nokinside 7y ago

It seems to be completely pie in the sky type of thinking. The only rational -non delusional- reason I see for the talk is if Musk does it to get better financing terms before the self driving hype dries down. I think taxi services without driver are coming but the the secret sauce is remote human operator working together with high level of automation. Driver is the main cost of taxi service. If you get one operator to remotely oversee 20 highly autonomous vehicles and provide input when needed, the cost of taxi-service drops into a fraction of what it used to be and service is better than fully autonomous driving.

izzydata 7y ago
I would never trust such a ridiculous system. I don't know that many other people would either. Although I would also not trust a fully autonomous vehicle either until more evidence is provided that we were even within 20 years of it.
notfromhere 7y ago
So humans that can barely drive a single vehicle will now be responsible for monitoring 20 vehicles in highly different environments that change at a moments notice?
- Nokinside 7y ago
  Not at a moment's notice. The automation must react during emergencies, but humans can navigate and solve stuff that the automation can't.
  - marsokod 7y ago
    Yes. The car is responsible for the safety in all circumstances and recognizes it's limitations. Think level 4 autonomy. As soon as the car sees it cannot proceed properly, it parks somewhere or go in a safe situation and requests a remote human operator to intervene. Then the whole point is to have a call center sized so that you do not have to wait too much. We're not there yet, and giving back the control to the passengers is also a possibility. However, in a world where passengers almost never drive, it might be safer to give the control to trained remote operators.

ajmarsh 7y ago

I don't think it's that far off at all. https://x-matik.com/ https://x-matik.com/

mruts 7y ago

Is it fair to ask at this point if Elon Musk is going crazy? His decisions of late seem unwise at best, irrational at worst. Maybe he should just concentrate on producing Model 3's and then the Model Y? Or what would be really cool would if Tesla could produce multiple car models... at the same time!

elorant 7y ago
More like desperate. He has to hit targets that seem impossible otherwise he'll run out of cash.

DennisP 7y ago

I've always been skeptical of skipping LIDAR, but a recent study at Cornell found that with binocular vision they could be almost as accurate as LIDAR, but with much cheaper hardware. https://www.therobotreport.com/researchers-back-teslas-non-lidar-approach-to-self-driving-cars/ https://www.therobotreport.com/researchers-back-teslas-non-l...

Judgmentality 7y ago
That study has been highly disputed by people in the field; the methodology is questionable at best for the claims they're making.
- toomuchtodo 7y ago
  Anthony Levandowski of Waymo LIDAR fame has said he now agrees with Elon Musk ("a wiser person than me") that LIDAR is a crutch. https://youtu.be/fNgEG5rCav4?t=105 https://youtu.be/fNgEG5rCav4?t=105
  - xmprt 7y ago
    I wonder why the person who was taken to court over (allegedly) stealing Waymo's IP related to LIDAR would be wary of using LIDAR in his new projects...
  - Judgmentality 7y ago
    Anthony Levandowski is so sleazy he makes Vinod Khosla look like a good guy. He is (well, was) one of the most untrustworthy people in the Valley, so much so he had to leave for China to start his next company because even local VCs - who care about money and nothing but money - found him too sketchy to work with.
    - toomuchtodo 7y ago
      You can be sleazy and untrustworthy and still be smart and right. I wouldn't use VC funding as signaling, considering how many VC folks blindly funded Theranos (which wasn't just outright fraud, but abusive pursuit of whistleblowers). It's fairly clear machine vision is going to surpass LIDAR in the near future (taking capability and cost into account), doesn't take a robotics expert to see that.

agumonkey 7y ago

You don't buy his claims that lidar is non necessary and even inefficient at sdv ?

mplewis 7y ago
Elon/Tesla have a vested interest in making sure LIDAR isn't required by law for self-driving vehicles. In addition, his NN-camera-powered-cars are still trying to drive themselves into highway barriers.
- Robotbeat 7y ago
  Requiring a specific technical solution with problems of its own by law is a wet dream for whoever holds the IP to that technical solution. Now accusing others of having a vested interest in MAYBE opposing this hypothetical law that no one has actually proposed is just gas-lighting and possibly projection.
  - notfromhere 7y ago
    LiDAR means he'd have to retrofit vehicles, given how he promised every purchaser that their cars are self-driving ready. LiDAR literally will cost him money, so yeah, that's a conflict.
- agumonkey 7y ago
  Yeah but back to its point, he argued lidar was not optimal for "computer vision", do you think it has limits that would be shadowed by multi camera NN analysis ? That he's doing bad PR is another point (that we can discuss also but is harder to comment on objectively)
- manicdee 7y ago
  Do you believe lidar is so necessary that it must be mandated by law?

senectus1 7y ago

He's not the only one working on this. I think its seriously deliverable.