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

Uber Wasted $2.5B on Self-Driving Cars

theinformation.com

Post reply on HN
Print view View original

171–180 of 231 posts

Re: Uber Wasted $2.5B on Self-Driving Cars

#171

Post by treeman79 »

ip26 wrote: post #132
xiaolingxiao wrote: post #45

Earlier quoted context omitted.

Indeed amazing because I know employees at Tesla who take it home on autopilot in the freeway! It’s also very unsafe ans I dot believe camera only is the way to go, but at least it can drive for more than a mile.

At this point I'm starting to be willing to consider it fairly safe when the driver is paying attention. Every press article about an autopilot crash seems to involve somebody who is completely inattentive. I can't recall any published incidents involving a driver using it as an assist, while continuing to pay attention.

They used to build very straight roads in the desert. People kept falling asleep at the wheel.

So they started Adding curves in the road. Not much, but it improves the engagement factor. Gives the person something to do.

It sounds horrible to have to be fully alert, yet nothing to do.

In programming, I can’t stand shadowing people for more then few minutes before I want to start driving the code myself.

Re: Uber Wasted $2.5B on Self-Driving Cars

#172

Post by rrrrrrrrrrrryan »

xvilka wrote: post #97
jimrandomh wrote: post #76

Earlier quoted context omitted.

Not really. Rail in the US is spectacularly expensive; look up a few recent rail projects with dollar figures attached, they're billions of dollars each. That's the same order of magnitude as the cost of an entire self-driving research program (Waymo has taken $3B of investment, according to its Wikipedia page), which, if successful, would scale quickly and easily to the entire world.

If the US companies can't do it just hire Chinese or Japanese ones who have enormous experience to build the infrastructure way cheaper.

The expensive part is addressing all the local regulations and acquiring land.

Human rights (and property rights) can be expensive.

Re: Uber Wasted $2.5B on Self-Driving Cars

#173

Post by abrichr »

phire wrote: post #93
wongarsu wrote: post #37

Earlier quoted context omitted.

Waymo seems to be doing much better than Uber, so it can't be just about what's possible. Waymo are also doing much more than just deep learning. For example a deep learning algorithm might be good for predicting possible paths of tracked pedestrians, but figuring out whether you will intersect any of those paths is better solved by a different non-learning algorithm.

Waymo almost predates the current deep learning fad. While I'm sure they have deep learning somewhere in the stack, the core of their algorithm appears to be based on more traditional techniques, with kalman filters being the core.

> the core of their algorithm appears to be based on more traditional techniques, with kalman filters being the core.

Interesting! Can you please share a source so that we can learn more?

Re: Uber Wasted $2.5B on Self-Driving Cars

#174

Post by treeman79 »

smachiz wrote: post #165
ghaff wrote: post #142

Earlier quoted context omitted.

Everyone has made disappointingly little progress on full self-driving in the sense of driving little Johnny from the house to soccer practice by himself. There are possibly opportunities on freeways but anyone who expects to be able to get robo-Ubers door-to-door for decades is likely to be disappointed.

lol. You're VASTLY underestimating how much progress is being made by companies that aren't Uber. 3-10 years is a more realistic timetable than decades. "People overestimate what they can accomplish in one year, and underestimate what they can accomplish in 10".

20 dollars it’s more than 20 years.

After all, writing an AI is just a summer project for couple of interns.

Re: Uber Wasted $2.5B on Self-Driving Cars

#175

Post by abrichr »

FandangoRanger wrote: post #20

There's just no way self-driving cars can meet the challenge with today's technology. These days the tests they are doing involve burying a wire in the road that the car can ride upon, which to me seems like covert recognition that computer vision and sensor data feeding AI is not equipped to handle the rigors of road travel quite yet. The theoretical arguments about how self-driving cars are better than human driver…

> These days the tests they are doing involve burying a wire in the road that the car can ride upon

Interesting, can you please share a source?

Re: Uber Wasted $2.5B on Self-Driving Cars

#176

Post by rdw »

paxys wrote: post #135
rdw wrote: post #121

Earlier quoted context omitted.

https://www.inc.com/jason-aten/uber-lyft-are-betting-that-dr... https://www.theguardian.com/technology/2019/apr/19/ubers-sel... https://www.bizjournals.com/sanfrancisco/news/2018/01/24/ube... https://www.latimes.com/business/la-fi-uber-ipo-stock-price-... https://www.cnbc.com/2020/01/28/ubers-self-driving-cars-are-... https://www.wired.com/story/bet-uber-bet-self-driving/ I think the broadest interpretation is that a…

Okay, so in the extreme case Uber's self driving division is estimated to be worth ~$7 billion? That proves my point more than it refutes it. Their stock has fluctuated more than that this past month.

That $7 bn valuation came only a year after the disastrous 2018 crash which forced them to shut down the entire program for months, with months of of lurid news stories highlighting failure after failure. Probably a low point, actually.

The developing narrative seems to be that self-driving was a longshot that could have paid off hugely, had it worked out, but it's been plagued with problems and it just seems very unlikely to pay off now. But still not zero, since no one else has shipped anything.

Investors probably valued it like (0.7% chance of success * 1 trillion upside = 7 billion).

Re: Uber Wasted $2.5B on Self-Driving Cars

#177

Post by partingshots »

FartyMcFarter wrote: post #38
partingshots wrote: post #34

Earlier quoted context omitted.

> Didn't they lose a lawsuit against Waymo, which meant they had to build a lot of tech from scratch and, to add insult to injury, keep Waymo informed of the details? That’s really interesting. Does anyone have a link to where this was stated?

Apologies, I deleted that last part of my comment after I tried to search for this and failed to find evidence. I'm not sure if I'm remembering wrong or if this detail was in some obscure article somewhere. The settlement is discussed for example here, but no mention is made of Uber having to disclose future developments: https://9to5google.com/2018/02/09/alphabet-waymo-uber-trial-...

No worries. I also recall vaguely at least, that they had to give up their existing work that had been based off of stealing from Waymo.

I don’t remember how exactly they were going to enforce things though, which is why I asked.

Re: Uber Wasted $2.5B on Self-Driving Cars

#178

Post by Gibbon1 »

yholio wrote: post #30

Clearly, we should aim for a less ambitious goal: ultra-compact self driving vehicles on dedicated and physically separated lanes, where cars cannot be encountered unless it's their fault, similar to bike and tram lines. This could increase city trafic capacity by an order of magnitude in congested central zones and suplement public transport in suburban, dispersed areas that cannot be properly served by buses. And i…

Reminds me of the The Morgantown Personal Rapid Transit system.

https://www.youtube.com/watch?v=iaSaWfw07Sw

If you used battery electric cars would likely be a lot cheaper to build.

Speaking of electric, in Taiwan they have occasional dedicated paths for scooters. With the advent of battery power bicycles grade separated bike lanes seem more useful. Because battery power eliminates the hill problem.

Re: Uber Wasted $2.5B on Self-Driving Cars

#179

Post by scoopertrooper »

jcranmer wrote: post #169
ethbr0 wrote: post #87

Earlier quoted context omitted.

I'd offer Kurzweil's (I believe?) observation that the futurist predictions that come true are those that substantially improve peoples' needs. Even substantially expanded rail will never solve the problem of "I want to get from point A to point B, as quickly as possible, whenever I'm ready to leave" in a fundamentally different way. The transit to the starting station might be shorter, and the transit from the endin…

When you pack people into a train, you fit about 1 person per m². Pack people into a single-occupancy cars, you get about 1 person per 10 m², or 0.1 person per m². That's a 10× capacity loss that has to be made up somehow if you want to move people as efficiently as possible. Autonomous cars don't solve the problem of inefficient use of land for transportation (actually, they probably make it worse since there will b…

A purely autonomous road system would make much better use of space than the current system. For instance, you could discard with the notion of fixed lanes on a motorway as the system could nearly instantly allocate lanes on the basis of demand. You'd also be able to do away with curb side parking in urban centres, which would free up more land for commuting.

A train is obviously more still more efficient between fixed points, but self-driving cars are far more economical for less-frequented routes and more convenient for last-mile transit than buses.

Re: Uber Wasted $2.5B on Self-Driving Cars

#180

Post by ericmay »

wmf wrote: post #156
ericmay wrote: post #126

Earlier quoted context omitted.

It blows my mind to see people write things like this. We have two companies, Cruise and Waymo, who have built nothing but toy cars, but they are in the lead while Tesla has years of real-world driving data in cars that are in a manufacturing production operation, receive over the air updates, recognize real world objects, can change lanes automatically, but Waymo is ahead? I’m open to being convinced that I’m not co…

Tesla didn't support stop signs or traffic lights until a few months ago but Waymo did that years ago. And AFAIK Tesla still can't recognize stopped vehicles on the highway and has no public plan to fix that.

Waymo hasn’t produce a single production car. I doubt their software works on random stop signs in Cleveland. Until I see them on the road, I’m ok watching incremental improvement happen. Being first doesn’t matter.

> And AFAIK Tesla still can't recognize stopped vehicles on the highway and has no public plan to fix that.

Can you elaborate on what you mean by this? I do have a Tesla (without full self-driving) and it comes to a complete stop when cars in front of me stop on the highway.

Post reply on HN