Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Waymo CEO dismisses Tesla self-driving plan: “This is not how it works”

arstechnica.com

Post reply on HN
Print view View original

401–410 of 581 posts

Re: Waymo CEO dismisses Tesla self-driving plan: “This is not how it works”

#401

Post by hobofan »

epistasis wrote: post #221
cromwellian wrote: post #194

Watching this video of regular folk (not tech journalists or other "reviewers", or paid PR) riding in FSD Waymo has done a lot to reduce my "but it'll never happen" feeling. https://www.youtube.com/watch?v=qAZ6tJSj9T4 It may very well be that future FSD modes will require highly detailed maps, manual curation, limited areas where things can work, but even that would be valuable. e.g. if self driving trucks only work…

Huh, I'm surprised you find those videos convincing about the future of self driving cars. To me, they don't seem much better than the demos people were giving ten years ago, in academic competitions, etc. The Tesla videos scare the hell out of me. My own auto pilot purchase scares me whenever I use it. I can't even use cruise control in my Tesla without it randomly slamming on the brakes. It's pretty straightforward…

> To me, they don't seem much better than the demos people were giving ten years ago, in academic competitions, etc.

From what I remember from ~10 years ago (and a quick glance at Youtube videos from then confirms it) the state back then was roughly:

- DARPA challenges where cars were driving on dirt roads at about walking speed

- High speed driving on completely closed of racing tracks (with no other cars present), e.g. BMW i330

- Simulated real world conditions on closed of tracks with the system significantly slowing down at intersections and pretty reliably producing slow speed crashes

Compared to what that video is showing, that's night and day.

Re: Waymo CEO dismisses Tesla self-driving plan: “This is not how it works”

#402

Post by ra7 »

xedeon wrote: post #399
ra7 wrote: post #392

Earlier quoted context omitted.

But that’s not what the GP comment is talking about, no? They’re specifically talking about self driving on highways. FSD beta works on city streets. On highways you’d use auto steer and traffic aware cruise control, which is exactly same as lane keep assist and adaptive cruise control in a Toyota.

Can the Toyota Safety Sense system perform an automatic lane change on the highway? I don't think it can. Please correct me if I am wrong.

I’m not sure, but I don’t believe it can.

Re: Waymo CEO dismisses Tesla self-driving plan: “This is not how it works”

#403

Post by ganafagol »

snoshy wrote: post #181
mxschumacher wrote: post #54

Earlier quoted context omitted.

it's interesting how optimistic long term visions conflict - if you really think autonomous cabs will become widespread then by implication you are also saying that car sales will be drastically lower. Tough to be serious betting on both. The math of "you'll pay more for the car but the car will go out and make money for you as a robotic Uber" sounds like a pyramid scheme to me; if too many people do it, it can no lo…

I disagree, largely because you've cherry-picked parts of the overall strategy. The world is obviously not going to go from purely human driven to 100% autonomous overnight. There will be an intermediate phase where initially some few cars will be autonomous and privately owned, and for those owners making use of the car's full lifetime to serve as a taxi would be a no-brainer. It could pay off your (very expensive)…

> It could pay off your (very expensive) Tesla in the matter of a year or two.

.. while lota of people unknown to you sit with their butt in your precious expensive car. People who themselves can't (or want to) afford such a car.

Not gonna happen. Nobody will throw that much money behind such an expensive gadget, being happ about their precious toy and then let complete strangers user it most of the time.

Do you rent out your home via Airbnb when you are 2 weeks on vacation? Very few people actually do that.

Re: Waymo CEO dismisses Tesla self-driving plan: “This is not how it works”

#404

Post by rcxdude »

Udik wrote: post #77
rcxdude wrote: post #34

Earlier quoted context omitted.

The core point is you can't go smoothly from a driver assist to fully autonomous. They're just completely different from the point of view of the reliability of the software you need. A huge amount of things with driver assist you can write off as 'the driver will take over'. There's not a specific part of you can point to and say 'that's what will stop them', so much as the entire design of the system needs a differ…

> you can't go smoothly from a driver assist to fully autonomous Ok. So Waymo has developed a 100% reliable full self driving system that handles every possible situation. It was never put in a situation in which it would have given back control to a human driver, because that would have meant "going smoothly" to driver assist mode, and we have already said that this is impossible.

I was not referring to the dynamic shift while in use, I was referring to the development process. Regardless, the waymo system does not have a driver assist mode, it's either in control or it isn't (and during development it may be overridden by or automatically fall back to a safety driver, who has a job akin to a driving instructor, something which general requires more qualifications and attentiveness than a driver).

Re: Waymo CEO dismisses Tesla self-driving plan: “This is not how it works”

#405

Post by FrojoS »

roenxi wrote: post #370
meekrohprocess wrote: post #151

Earlier quoted context omitted.

I have one question for Tesla customers who trust the company to deliver full FSD. How do you reconcile that belief with the fact that Tesla's embedded developers did not understand the extremely simple concept of write endurance? The NHTSA opened an investigation into premature HUD failures because they prevented the backup cameras from working. But the fact of the matter is, the company used a small partition of in…

There is no challenge reconciling imperfect FSD with high trust of that FSD. When I decided to play the game of risk minimisation, I sold my car. Minimising risk isn't the most important goal of drivers, almost by definition. Cars are not safe in any objective sense. They are tools of convenience. A fun hypothetical, you and a good friend get tested for spatial intelligence and it turns out there was a big difference…

Cars are most dangerous for pedestrians, cyclists and bikers.

Re: Waymo CEO dismisses Tesla self-driving plan: “This is not how it works”

#406

Post by rcxdude »

_ea1k wrote: post #180
rcxdude wrote: post #37

Earlier quoted context omitted.

The whole point is they aren't actually doing what waymo is setting out to do, and it doesn't matter how many cars tesla has with driver assist it won't really help them get to full autonomy.

If having a lot of cars with sensors collecting data doesn't help, why did Waymo build so many?

Because testing the cars does help. Beyond a certain point you also need information on how the system responds, and not just in the areas your customers are comfortable having it engaged.

Re: Waymo CEO dismisses Tesla self-driving plan: “This is not how it works”

#407

Post by elisharobinson »

rcxdude wrote: post #46
screye wrote: post #35

I worked with a self-driving product team a couple of years ago. Tesla's approach was looked upon with suspicion and often openly mocked. Tesla's entire self-driving strategy is led by Karpathy, who while brilliant, is an under-experienced researcher with a narrow expertise in large scale 2D CNNs. Tesla (and Comma.ai ?) seem to be the only groups that find Lidar unnecessary for L5 self-driving. I completely agree wit…

This matches my experience of how Tesla is viewed in the autonomous car industry. No-one thinks they are a serious contender when it comes to a driverless vehicle (and their business model basically doesn't match with the technology either: every other company is basically going to operate a taxi service because there's far more value to be captured and it can help pay for the hardware making driverless possible)

Really though how realistic is your business model ... if you wanted a autonomous network to replace Uber Currently 30 million cars on road . you would atleast need 10 million cars assuming each car costs 100k ( optimistically with lidar tech factoring in capex for factories and you dont have major product revisions ) thats 1 trillion in capex ... now you if you had the revenue of apple in 2020 ~ 0.25T$ you could expect to be cashflow positive in 7~8 years (assuming exponential ramp up of production and 0 competition) ... all of these thing need to happen for this taxi service X to succeed ,i think to get all these things right is highly unlikely and i feel Elon knows this aswell so he took the path of simpler tech and trying a software service model which is infinitely more scale-able .

Re: Waymo CEO dismisses Tesla self-driving plan: “This is not how it works”

#408

Post by ra7 »

mgoetzke wrote: post #391
ra7 wrote: post #26

I think the FSD beta rollout is completely irresponsible from Tesla. Rolling out a clearly half-baked safety critical technology to its customers (yes, I know it's only a few beta testers) who are untrained is nothing but a tactic to generate hype and get more customers to buy that $10k FSD package. This is on top of Tesla being the least transparent company out there in reporting safety data or their testing methodo…

That safety 'Report' from Waymo isn't a report at all though. It is a general overview on their approach how they say they will handle it and a direction. Basically a primer for regulators and customers that don't know anything yet about it. It is extremely light on numbers over time.

There are a couple more PDFs linked in the same page. What did you find lacking in the safety performance data whitepaper?

Re: Waymo CEO dismisses Tesla self-driving plan: “This is not how it works”

#409

Post by snovv_crash »

KorematsuFred wrote: post #315
NationalPark wrote: post #91

Earlier quoted context omitted.

Surely you use power steering, drive-by-wire pedals, and anti-lock breaks though, right? Or even an automatic transmission? Is it so different to use lane assist or cruise control?

None of them have to deal with fat tail risks. (to quote nissim ntaleb). I do not know about dbw pedals but power steering and anti-lock breaking operate 100% deterministically and the only true unknown is controlled by the driver.

There's still mechanical failure to deal with.

Re: Waymo CEO dismisses Tesla self-driving plan: “This is not how it works”

#410

Post by JonathanFly »

screye wrote: post #35

I worked with a self-driving product team a couple of years ago. Tesla's approach was looked upon with suspicion and often openly mocked. Tesla's entire self-driving strategy is led by Karpathy, who while brilliant, is an under-experienced researcher with a narrow expertise in large scale 2D CNNs. Tesla (and Comma.ai ?) seem to be the only groups that find Lidar unnecessary for L5 self-driving. I completely agree wit…

>However, in the unlikely scenario that 2D videos are sufficient for L5 self-driving, Tesla wins BIG. No one will be even close, especially due to the scale of 2D data that they have gathered.

Potentially Tesla's 2D data lead means they will still be the leader even if the compute resources in the current cars isn't up to the task, and maybe even if they have to add additional sensors like lidar. They can probably retrain their models and still get a lot of out of it.

My personal guess is that full 3D lidar isn't needed, but some additional cameras and other cheap sensors probably are. There's a ton of stuff that costs a fraction of the cost of lidar that might fill in the gaps.

Post reply on HN