Baidu’s self-driving cars begin public test in Wuzhen, China | Hacker News Reader

Baidu’s self-driving cars begin public test in Wuzhen, China

techcrunch.com

134 pointsbecewumuy125 comments

There is so much talk in the current rhetoric about the US losing its dominance in science and technology. But it is a tad myopic to not be excited about a country of a billion people pushing the boundaries of what is possible. It is not a zero sum game and we will all benefit.

And a bit of healthy competition for national kudos is likely to speed things along.

I agree. Just as long as we don't let safety fall by the wayside in order to speed progress (it has happened before, and it will happen again: see early nuclear refinement).

How? How do the uneducated in the rust belt benefit? And perhaps more importantly when?

I agree that technology for society in general make us better of over time, but society consists of individuals who don't necessarily benefit from this. In fact 12 million jobs are in danger of being lost to automated cars. How are they going to benefit from that?

Well the current generation will suffer, but the next generation and future ones will benefit from self driving cars.

Who owns the future though?

While I am not a nationalist and I think it's commendable that the human race is making progress, I'm concerned that advances in technology are causing troubling concentrations of wealth among those who operate and control the technology.

luddites thought the same ;)

The luddite fallacy is in itself a fallacy as we are today not just talking about replacing physical labour with technology we are also replacing jobs that require mental abilities and at higher and higher levels of abstraction in an environment thats exponentially progressing.

The cold hard reality is that a lot of the jobs that's been created in the US are wallmart jobs or part time jobs and that the cost of living is going up.

So those who keep mentioning luddite fallacy owe it to explain what new jobs it is they see replacing the current jobs. Because if it's just a hope that it will be all right in the future some day it's not really an argument for anything.

Luddism was about a privileged caste defending their way of life. That has little to do with thinking about a post labor world and what happens to everyone who gets stranded outside the cash economy, never mind having any capital. If coders formed a guild and smashed automatic coding machines, now THAT would be more like Luddism.

What do you suggest though? Should we intentionally hold back technology so that factory workers and bus drivers have jobs? Especially as these automation schemes generally have the promise of drastically increased safety.

I agree it is a problem that people will be put out of work, but that is kind of an inevitability. Which is why a lot of people are arguing that basic income needs to be investigated much more closely.

Thomas Piketty has articulated a path forward. I support his view that governments ought to step in and institute a global tax on wealth.

Except that doesn't solve this problem (mostly: Read all the way through)

The issue isn't The Wealthy having all the money. The issue is technology that will actively remove jobs. It doesn't matter if minimum wage is 90 pounds an hour if you don't have a job. An automated train isn't The Wealthy stealing money. It is a conductor out of work because their job could be handled by a raspberry pi.

Basic Income is a theoretical concept (been implemented in small, borderline meaningless, experiments) to combat that. Everyone gets a bare minimum "salary" regardless of if they work. That covers cost of living (how much more varies). Then, those with jobs have the potential to earn a lot more (incentive to work) on top of that.

What Thomas Piketty proposes wouldn't address the lack of income to those put out of work by technology but could, possibly, help to pay for solutions that would.

I would say that the issue is The Wealthy having all the money.

Take it a few steps forward. At some point the factory owner (whether local or conglomerate) holds all the money and has 0 workers. The only thing he can buy is new resources to produce more stuff, but no one can buy that stuff because no one works anymore.

At some point it's just a machine making stuff that no one can afford to use other than the very few owning the machine.

Money is useless unless it's spread around and in rotation through the system. The unemployment of uneducated is the one thing I really sympethize with in this election, but the solution isn't going backwards, because there's nothing there anymore.

I understand that the solution isn't one step forward, but maybe 5, and people will suffer in the interim, but that only means it's ever more important to swiftly move in the right direction.

correct. If the only way to have money to spend is to work then if nobody works nobody can buy anything. UBI is one way to make people who don't necessarily work have money to spend. Taxing the wealthy is one way to fund UBI (the other is printing money, which is basically taxing savings indiscriminately by causing inflation). I honestly don't know if that would work on practice. But the point is that merely taxing the wealthy and do nothing to address the fact that people without jobs don't get money won't fix the problem of jobs disappearing. However there is a big social and psychology barrier that prevents giving money go those who "don't deserve it" by sweating for it. On of the big drawbacks in current welfare systems though is that whoever currently benefits from them has little incentive to work a little bit because they would get the same amount of money but they'll actually have to work. The only incentive is to get a job that would pay more than what you get from welfare and that creates a barrier and also reinforces the stereotype of some people just not be willing to work (e.g. because of laziness).

Well if that happens, then the population would shrink to accommodate (and it sort of is happening as people are having fewer children).

> In fact 12 million jobs are in danger of being lost to automated cars. How are they going to benefit from that?

Jobs are going to be lost anyways, whether from cars or from a different technology.

Are you arguing the problem is that people aren't able to continue to do trivial things which technology can easily replace them at? Or that they will not have a relevant skill / source of income?

If the former than we should give hoes and scythes back to 90% of the population and get rid of agricultural machinery. If the later - it's a problem that needs to be addressed one way or another, self-driving cars are really a minor detail here..

I was responding to the OP who said it wasn't a zero zum game.

Not sure why we are suddenly discussing what I propose instead.

All I am saying is that it's easy to forget that real people are effected by progress and that the narrative that we all benefit is kind of disingenuous from that perspective and I believe we need to think harder on what to do about that because it's only going to increase.

That was a quick backpedal. :)

In what way?

Yep, agreed. Universal basic income is one solution that keeps coming back as a silver bullet. Not sure what the downsides are yet, but I'm pretty certain that there's still going to be haves and have nots: just a few rich that own all the automation and resources vs the rest of us just surviving on basic income.

Maybe we'll all be ok immersed in VR adventures as long as we're fed and clothed...

One danger is that it lowers the incentive to work.

You will always be incenified to work when it's basic income at least the UBI version as it's not the negative tax version that some talk about.

If there is little need for work, extrinsic incentives to work don't matter much. Intrinsic incentives (i.e. work is fulfilling) will be enough.

But that often isn't enough to get people to work on boring but necessary work.

> boring but necessary work

Sounds like a fantastic job for machines! Humor aside, even with UBI you will always increase your income and wealth if you put in extra work, unless some current schemes where your income stagnates or even falls if you earn enough to put yourself outside some arbitrary bracket.

>Not sure what the downsides are yet

It's fantastically expensive.

As a share of the USA's Gross Domestic Product, health spending accounted for 17.5 percent. Single payer systems like Canada are 10.9% and have similar or better health outcomes.

So, ~4.4% GDP is pure healthcare waste. Social Security is 4.9% GDP and welfare is another ~4%. So, right now we could use 13.3% GDP or 7k per person in the US. However, that 7k would remove the justification for a lot of low income tax breaks freeing up enough to hit the 1,000$ per month range without raising net taxes.

Now, this would take a huge shift at the federal and state levels, but compared to say the 'New Deal' it's not actually that large a change in policy. Further, while some people would stay at home, it would also free up a lot of jobs and inefficiency.

The real problem is a small but very influential slice of the population would be worse off.

PS: Remember, every policy change is going to make somebody worse off, so harming someone is not enough to make a change bad.

Not any more expensive than paying all the wages under the current system. Besides, it gets rid of the current welfare system's eligibility-determining bureaucracy, so it might even be cheaper.

For sustaining the basic income the owners of the automation and resources should pay taxes. At least a bit. How much tax do the big players pay?

It's politically infeasible. You might as well ask "why doesn't everyone just get along? what's the downside?"

People seem to be too focused on jobs that will be lost because of automated cars/vehicles, but are not looking at the jobs that will be created because of that.

Think back to when people were just becoming aware of computers and how they will change the workplace. The same arguments were made - high unemployment, etc.

Look where we are today. Imagine a world without computers!

I see the same for automated driving. It will lead to the creation of more jobs than it will kill. Yes, some jobs will go, just as computers killed some professions.

Get ready for change folks. Embrace or go the way of the dodo

What jobs are you talking about? That's the question. Those jobs you claim have come instead, where are they? The computer revolution is decades old now.

You must be able to point to some jobs that actually provide more, just claiming it will create more while the reality is that it doesn't unless we are talking wallmart jobs isn't an argument thats more a religious belief.

There are around 3-4 million software developers in the US. 18million worldwide. Those and some other areas of STEM jobs are providing some of us with a future and rising income. But for most other people they are on the wrong side of that divide.

What jobs are you talking about? That's the question. Those jobs you claim have come instead, where are they? The computer revolution is decades old now.

So you were not old enough to be part of that discussion.

But what jobs am I talking about?

Back then, the counter argument was computers will create jobs for those who have to design them and those who have to maintain them. That was the best they could muster. Very little was said about computer networking back then.

But today, we have developers, system and network admins, operations people. And that's just those that are directly involved in managing computers.

So don't panic. Robots, autonomous vehicles will create more jobs than they'll kill. Humans will be needed to build and maintain then. In many instances, they will need to be networked. Humans will be needed to set up and maintain such network of robots and fleets of autonomous vehicles.

The future is bright. Embrace it!

I am 42 so I've been around and you aren't answering the question which is telling.

There are 18 million developers world wide, a little more than 3million in the US.

Between 1998-2004 the us lost 4 million jobs to china, in the same period china lost 15 million jobs to the robots.

Outsourcing is just the step before automation.

I am not panicking as I am one of those who benefit mostly from this, I am just painfully aware of some of the consequences. And you haven't provided a single example of new jobs for those without education which is the base of this discussion.

Let me have a go:

When I was at university google had 8 employees :) ... who would have imagined then jobs like SEO/SEM Consultant, blogger, online life-coach, android indie game designer, stay-at-home person selling junk on ebay, etc, etc.

For self driving cars... how about:

- Guide (jump into the self-driving van, and I (human) will give you a personal and engaging tour of the region, parks, people, wines .... we'll do a few stops at designated restaurants and souvenir stores (where I'll get a few dollars per visitor).

- Self-driven drug / alcohol / party stuff / shopping / pet / kids / delivery

- Artist (drawing random patterns on the dusty planes with programmed cars?)

- Game host (where you offer 50 cars to some players who do some kind of AR laser tag activity while driven ... all the while trying to figure out the patterns they are driving in to win)

- Real-life marketing consultant (sell pay-per-detour, is paying for the detour worth it? Only if they used keyword X last, week and you have the right store front design, blah, blah, blah).

- Scheduling Assistant (Human helper to call your rides)

- Mobile Mechanic (fast response team to fix those helpless cars stopped on the road because a leaf covered the LIDAR)

Will the number of jobs add up? It's anybody's guess. I think the transition is going to hurt badly. But it will reach a new equilibrium eventually.

You don't have to guess the numbers wont add up. Google and Facebook have way less people than other companies their revenue taking into consideration.

What is happening these days is an increase of capital rather than labour heavy companies. The number of new companies is falling in all of europe and the size of the large companies is falling net.

Most of those things are on the verge to be automated too.

So you are not even remotely showing anything.

When you say "Google and Facebook", you are talking about their employees. But how many new companies / jobs have been created using their ecosystem (How much additional sales and jobs are created with those platforms?). That number is a lot harder to establish.

Do you have any estimates or references about _that_ number?

> finid 3 hours ago | parent | on: Baidu’s self-driving cars begin public test in Wuz...

People seem to be too focused on jobs that will be lost because of automated cars/vehicles, but are not looking at the jobs that will be created because of that. Think back to when people were just becoming aware of computers and how they will change the workplace. The same arguments were made - high unemployment, etc. Look where we are today. Imagine a world without computers! I see the same for automated driving. It will lead to the creation of more jobs than it will kill.

Not sure how you figure that.

I'm pro self-driving cars.

But you have to admit, there's going to be riots in the streets when it's no longer economical to pay a human to drive a trunk/taxi.

Look at the protest people have over Uber, and they're not even destroying jobs, they're just moving them from one place to another. And already you can see that non-trivial parts of society can't handle that.

There's going to be protests, poverty and chaos, and eventually the market will redistribute the workforce but short of a major welfare program like 'basic income' I don't think we can have any illusion of this being a soft transition.

> But you have to admit, there's going to be riots in the streets when it's no longer economical to pay a human to drive a trunk/taxi.

Why would there be riots in the streets for self-driving cars in particular? Like when there were riots for the assembly-line or the tractor? There will be riots for education, healthcare and opportunity - and probably there should be. Technological progress of civilization should be something more meaningful than the rich getting richer.

>Like when there were riots for the assembly-line or the tractor?

Yes, exactly like that.

https://en.wikipedia.org/wiki/Power-loom_riots

https://en.wikipedia.org/wiki/Swing_Riots

People used to have jobs walking behind horses in fields, while the horses pulled ploughs and so forth. Pretty stultifying, dawn to dusk staring at horse's ass. A generation from now, manual driving will be viewed as maddeningly dull makework for human drones. Hours of staring at roads with instant death the punishment for dozing off? Wouldn't do it if you paid me.

I'd have said, hopefully the world gets basic income and then a full leisure society sorted out. But honestly, not feeling optimistic right now. The Trumpoids and Brexiteers voted to get their makework back. They won't get it, the economics will make sure of that. But they probably will get unemployed.

How many of the horses have jobs today though? :) But sure I agree. That doesen't change that for some people their job is all they have. If you take that away they have no realistic future unless it's some post scarcity society.

"Canned horse meat was introduced in the United States under the Ken-L Ration brand after WWI as a means to dispose of deceased horses" -- Wikipedia

Are you suggesting a Soylent solution?

> That doesen't change that for some people their job is all they have. If you take that away they have no realistic future unless it's some post scarcity society.

This has always been the case throughout history: new technology makes some jobs irrelevant, and humans adapt by doing new jobs. It's unfortunate for people who cannot or will not adapt, but historically people have been resilient to such changes.

It's a legitimate issue when there aren't enough jobs to employ a significant portion of the population, and perhaps we're starting to reach that point now. We're seeing a decoupling in our economy: job growth is slowing while corporate growth and profits continue to rise. On one hand, this indicates that we are heading towards post-scarcity conditions. On the other hand, the current arrangement is that an increasing amount of wealth is being distributed to a decreasing number of people, leading to greater wealth inequality.

What can be done? We might increase wage/salary while cutting the number of hours worked per week, allowing more people to work while still providing a livable income. A basic income is another option, but that seems extreme; things would have to get much worse (or better, depending on your long-term view of automation) before that's taken seriously on a large scale.

It hasn't always been the case though.

It's only been the case for around 200 years. Before that the progress was so slow that there was no issues transitioning as most kind of work took generations to change.

The problem is that now things changes so fast that lots of people simply can't re-educate them selves and the market doesn't really need that many people, yet we have no plan what so ever for this issue besides UBI and a hope that technology will allow us to create a post scarcity society. Hourly wages rarely even make sense since thats exactly the kind of jobs that normally could be calculated that way which are going away.

I would really urge anyone who think that technology creates more jobs than it removes to show where those new jobs are besides to the countries we've been outsourcing them too.

But jobs moving to China and India isn't solving the underlying issue and I simply don't understand why people don't take it more seriously and why Luddite fallacy keeps coming up. It's not that good an explanation (not saying you talked about luddite fallacy just in general)

> But jobs moving to China and India isn't solving the underlying issue

I thought we were talking about automation and technology? What do China and India have to do with this? I never said it's good for a society when their jobs are displaced to another region.

> I would really urge anyone who think that technology creates more jobs than it removes to show where those new jobs are

I would really urge people to think of "jobs" and "productivity" as separate metrics, since productivity will likely continue to climb as the number of jobs continues to fall. The challenge will be reacting to that effectively; trying to reverse the hands of time to bring back jobs is futile.

If jobs are moving from one region to another, that's mostly unrelated to technological progress. Instead, it's because low skill labor has moved to low cost markets with low standards of living. We'll get these jobs back when we have a lower standard of living then rural China and India.

In the meantime, our scientific and technological edge has been one of the few things keeping our economy semi-competitive. Can you imagine if we lost the low wage jobs and the highly skilled jobs? We probably won't have to wait much longer.

I was about to answer you first question but I can see you removed it.

With regards to your second:

We are. I was just trying to pre-empt the claim that more jobs have been created.

Sorry, I removed it because I realized I misinterpreted your comment.

I am thinking about jobs and productivity as two separate metrics which is exactly why I am coming to the conclusions I am.

Productivity goes up because of technology but those who gain financially are mostly those who work in technology or own resources.

Cost of living is going up too which while salaries for most aren't. The kind of jobs you can get if you aren't in a few privileged industries are wallmart jobs or part time jobs. 12 million people are about to get a run for their money when the one job that couldn't be outsourced suddenly can to robots.

So no I don't believe the competitive edge in the US is really about it's technological or scientific edge but rather a host of other things like the dollar which allow US to stay afloat despite it's soon to be 20 trillion dollar deficit, the size of the market, it's access to the international markets, the system which doesn't redistribute well and so on.

I have no interest of low-wage jobs coming back. What I do have an interest in is that people recognize that a lot of people are being left behind not just brushing it aside as the price of progress. Because never before have so many been left behind with no real chance of a future unless it's on welfare. And if you think this election was bad just wait til the next one, unless we start recognizing the problems and actually address them.

The last time there was a huge technological upheaval -- the automation of farming -- there was a huge push for education that correlated with this, and that's what made a lot of the difference. Many of the state leaders in the high school movement of the 1910s and 1920s were farm belt states like Iowa and Nebraska. (http://www.theatlantic.com/magazine/archive/2012/07/the-triu...). That really made the transition from agrarian society to manufacturing a heck of a lot smoother.

Unfortunately I really don't see anything near the push for education and training today. And that's an issue. I'm sure there was plenty of pains in the agrarian transition, but education gave an out that allowed the next generation to prosper. Without a good education or vocational training type system, it's not just the current generation of adults that might have pains with the automation transition, it's their kids as well.

How does "basic income" work? I have never heard it explained in a way that made any sense. Starting with a premise that resources are finite, you will always have power brokers who oppose the idea of spreading resources thin across the population. To think that would ever happen on a global scale seems as likely as us populating mars. It could happen, but certainly not in my lifetime.

I admit I haven't given it a whole lot of thought, but, for the US at least, I see it happening slowly over time as a smooth extension of welfare, not as some piece of landmark legislation. As automation increases and jobs become more scarce, more people will need welfare assistance. With strained resources the government will look for solutions such as an automation tax for corporations, perhaps less defense/military spending, etc. Taken to it's logical conclusion, I think you end up with something like a basic universal income.

I see. That makes sense. I was attempting to understand it as some sort of, as you say, "landmark legislation". But picturing it more as an extension / evolution of welfare, definitely makes it seem more plausible. Though, welfare is something that you have to qualify for, and my understanding of basic income was that even the richest of the rich, would be able to collect their basic income.

Power brokers oppose all sorts of redistribution but public movements can demand it.

A minimum wage is already part way to a basic income. The other half, a minumum wage for just being human and alive, would require similar political support.

> hopefully the world gets basic income and then a full leisure society sorted out

Your job is gone? Here just have money.

Dignity? Oh no see you don't need that. Just have money. What will you do all day in a world where people derive a lot of purpose from jobs and your basic income isn't enough to just do what you want? ummm... we could legalize pot.

I'm a huge believer in the power of the individual, and under today's conditions I wouldn't support a universal basic income. But in a possible future where AI is sufficiently advanced to handle the majority of 'jobs' and where corporations dont employ people at the scale they do today, I would support a universal basic income. What would your solution be?

You think working two fast food jobs just to live is dignified?

Dignity? So you think economic coercion and slavery-with-a-choice-of-masters (or the non-choice of none, and starvation) is somehow dignified?

A leisure society is being able to do your vocation. It's about hackers being able to hack without needing a "day job" and without trying to squeeze what you want to do into the leftover energy after spending your day on makework.

I'd guess you got downvoted because that's irrelevant. The benefits come in the form of fewer road deaths, which is potentially as significant as curing malaria, which only in the abstract benefits the people you are talking about, but it nevertheless does benefit them.

Did you read the parent I am replying too?

How is road safety relevant to a discussion about zero sum games?

For one, it makes it a non-zero-sum game.

The claim was that we will all benefit. So I am asking how is that benefitting those who loos their jobs.

Sure you can cherry pick your perspective and make OPs statement true, but whats the point of that.

About a million people die on the roads each year. Even though the US has a relatively low rate of road fatalities, there are still tens of thousands of people killed each year and hundreds of thousands injured on the roads. Even the poor and jobless benefit from not getting mangled in a car wreck or run down while cycling, and the health care money saved can go to directly aid them.

What does that help them if they can't feed their family because they have no jobs.

> So I am asking how is that benefitting those who loos their jobs.

They get cheaper and better self-driving cars. I know that's not nearly enough to compensate for lost income and self-worth. The rest of society has easily enough gains to compensate them property, but because of outdated ideas about work, worth and capitalism we haven't done so.

But since they don't have a job they can't buy them but yeah.

Sounds like it's time to leave the rust belt then.

Well step 1 is accept that the jobs are gone/will be gone and they aren't coming back.

Step 2 is on the TBD. We'll let you know when we promise. Just around the corner really. We promise.

I'm serious about step 1 though. Should be preparing now.

Baidu's Autonomous Driving division is in California.

Self driving cars are much more valuable to China than the US. China, for all its development, lacks road and parking infrastructure and places to put them even if it was willing to build them. So cities like Beijing are incredibly congested. Likewise, the driving culture in many Chinese cities (all slightly different) is often very bad, it isn't uncommon to see someone back up on a free way because they missed their exit in Beijing....The US, for the most part, doesn't have these problems (lots of roads, parking, and fairly well behaved drivers)...self driving cars are seen as more of a convenience than a necessity.

Self driving cars potentially solve all that by optimizing limited infrastructure and standardizing driving behavior. It has the potential to make cities that are bursting at the seams (like Beijing) much more livable and viable. Couple that with an authoritarian government that can dictate rules much more easily, I predict that self driving cars will be a thing in China much faster than in the west (even if much of the tech still comes from the west).

it isn't uncommon to see someone back up on a free way because they missed their exit in Beijing....The US, for the most part, doesn't have these problems

I'm guessing that you haven't driven anywhere near New York City or Boston. I've seen my fair share of drivers backing up or driving on the part of the road they're not supposed to just because they missed their exit. Someone said I've done that myself, but I'll need to see proof.

Sure it "happens", but do you see it on almost all of your commutes? When someone does something dumb ass in the states, everyone is shocked and surprised, in Beijing it is just normal and should be expected (as a benefit, everyone is driving like everyone else is crazy).

(And I wasn't driving, rather sitting in the taxi, but I digress)

I live in the Seattle area and I see people doing "dumb ass" things on the road every single day.

I'm from Seattle originally. Every time I visited Seattle (well Bellevue/Redmond) for work, it wasn't that bad. People are pretty mellow, well, I guess you would have to go to Beijing to understand that there are different levels of dumb assery.

I'm originally from Brazil, so I've seen my fair share of traffic awfulness :) What I see in Bellevue is mild in comparison, but just as an example, today someone cut an entire lane in front of me to enter their garage.

An almost weekly occurrence is seeing people stick half their car onto the street when leaving a parking garage I drive by often, and only then stopping to watch for incoming traffic. I've parked in that garage before and I know there's plenty of visibility from the exit; it's like people don't realize there's a whole chunk of metal before the driver's seat.

> An almost weekly occurrence is seeing people stick half their car onto the street when leaving a parking garage I drive by often, and only then stopping to watch for incoming traffic. I've parked in that garage before and I know there's plenty of visibility from the exit; it's like people don't realize there's a whole chunk of metal before the driver's seat.

This would never happen in China! Yielding? Ha! Instead the car would just plow out of the parking garage irrespective of any cars or people on the road that weren't right in front of them! Same goes for right turns and anywhere else yield might be required...the concept just doesn't exist in China's driving culture.

Here's how bad China is.... I rode with a guy a couple of miles into incoming traffic just to impress me.... on a motorcycle.

Remember those Scifi chase scenes in movies like fifth element.... Felt like that.

Bad driving behavior exists everywhere, but in China it's much more common than in the US. Left turns from the extreme right lane (and vice versa), is another common thing you'll see. In my experience, the drivers in Beijing are better behaved than those in Qingdao. Lane changes without signals (while you're in the lane) are also really common.

Left turns from the extreme right lane (and vice versa)

New Yorkers in Manhattan know that dashing from extreme right to extreme left is the norm for yellow cab drivers.

<i> The US, for the most part, doesn't have these problems (lots of roads, parking, and fairly well behaved drivers)...self driving cars are seen as more of a convenience than a necessity. </i>

The same problems exist in metropolitan areas in US as well, like Bay Area, NYC, Chicago, etc. although to a less degree than those in China. In Bay Area, people are so exhausted by long-commute driving. Check how jammed the traffic is in 101, 880, 580, 237, etc. Bart and CalTrain's coverages are pretty limited. With self driving cars, I bet plenty of people will quit driving by themselves for the commute. Imagine one can read, relax, read a book, or even work in the car, instead of sitting behind the wheel for 2 hours a day.

Well, in china I just take a taxi. It wasn't a problem for me. But even though I had the convenience, capacity is an issue. I just moved to LA from Beijing, and I'm so impressed how nice the traffic is and how convenient parking is.

You don't see 11 day, 75 mile long traffic jams in the US though.

http://abcnews.go.com/International/chinas-traffic-jam-lasts...

To be fair, that was mostly inter city truck drivers who didn't want to pay tolls while they were reconstructing the only non toll alternative.

Your point is that because we don't have 11 day traffic jams, means we should be content? We're being unreasonable to want to hours of our lives every week?

I'd love it if all cars in China were self-driving. The road situation causes me stress whether I'm a pedestrian, cyclist or a car passenger (I don't drive here in Beijing). A few weeks ago, it took me over 3 hours to cover 16km in Beijing. It would have been faster to walk. The visible causes were:

- Drivers in Beijing drive slowly when there's heavy rain

- The sheer number of cars

- Drivers in Beijing advance into traffic-light controlled intersections even when the path ahead of them is blocked, causing a junction to get blocked up, even when all lanes leading away from it

I'm not sure what the other causes were (I heard that road space had just been reduced by the addition of a bus lane on the third ring road). I'm not sure what made that day much much worse than others (and it wasn't just me - I had a friend report a similar speed coming from a different part of the city).

I really wish Beijing had yellow boxes, and strong enforcement of not-stopping-in-junctions, like in London: https://tfl.gov.uk/modes/driving/red-routes/rules-of-red-rou....

I wish even more that the problem of humans needing to drive were solved.

But I'm not sure "Self driving cars are much more valuable to China than the US" is true. Each has a similar number of cars overall. China might skew more towards urban. But the cost for a full-time driver in China is 15%-25% of what it would be in the US, so the amount of GDP replaced by self-driving cars would be higher in the US, no?

Moreover, the lack of parking won't be solved by self-driving cars alone. People who can afford their own car will need to be persuaded to use car-sharing services instead. Without this, people will still want their cars to be near where they sleep, meaning that unused cars will still be jamming up the city whilst unused.

China lacks road infrastructure in dense cities, and lacks the room to build what it actually needs. By moving to self driving cars from say taxis, road usage can be increased dramatically (closer following distances, faster speeds, cooperative traffic control). It isn't the cost of the driver that is the issue here, it isn't the same benefit as we are expecting for the USA.

People in china are already ok with Uber, didi, and of course taxis. The private car was more about prestige and convenience (availability), not sure what to do about the former, but the latter is fixable easily enough.

In Europe, during the 18th and 19th century, the kings and emperors decided that they wanted grand esplanades and straight roads so they could have nice parades, preferably between some nice castles or parks. So they just demolished everything that was in the way. But hey! Marching in cool hats!

I am not sure about U.S, but my guess is that the cities started to grow a little bit later, the people probably looked at the European kings cool-looking roads, and probably decided that they could do even better and just went for an even cooler looking grid instead just some avenues scattered here and there!

- Drivers in Beijing advance into traffic-light controlled intersections even when the path ahead of them is blocked, causing a junction to get blocked up, even when all lanes leading away from it

This is the literal "gridlock" problem, which can lock up a road system. The US had that problem in some cities until New York City introduced "Don't Block the Box" around 1980. It took a PR campaign, and, some ticketing, and painting a big box with diagonal lines across some intersections, but eventually drivers got the idea to not enter the intersection until they had a way out of it. San Francisco is starting to have gridlock again, and a new "Don't Block the Box" campaign was started in 2014.

interesting. I wonder how automated traffic cameras/redlight ticket systems would impact that? I suspect most people "squeeze" into the intersection to avoid the next cycle, because they dont see a traffic cop around. if all intersections are photo enforced, you'd think that problem would go away (or at least, citations would go up enough to fund some better traffic control/flow solutions)

It's more a training/PR problem than an enforcement problem. Painting intersections prone to gridlock with a diagonally hashed box helps more than ticketing. Here's the Ministry of the Interior of Quatar trying to get through to drivers on Twitter.[1]

"In this business there are very few problems that can't be solved with some yellow paint and a little bit of common sense." - Henry Barnes, NYC traffic commissioner.

[1] https://twitter.com/moi_qataren/status/643829644253261824

China (well, Beijing) has tons of photo enforced intersections, but the government itself doesn't see "blocking intersection" as wrong so they never ticket for that.

Also, I've never seen a traffic cop actually pull someone over or do enforcement at all (maybe some DUI checkpoints, that's it!). If it wasn't for the cameras, no laws would ever get enforced. They only enforce parking laws a couple of times a year, so many two lane roads have become one lane as citizens take advantage of a lack of parking enforcement and a lack of cheap parking.

And yet every now and then I'll be stopped waiting for a blocked intersection to clear before advancing, and the asshole behind me will start honking.

I live in the Seattle area though, not NYC.

If the authoritarian nature of the government allows better enforcement, why are they so terrible at enforcing existing rules?

They focus authoritarianism on preservation and enrichment of the ruling class, everything else has less priority.

1.4 billion :)

I worry about the regulatory aspects of getting self-driving cars off-the-ground in the United States. It is inevitable that self driving cars will make mistakes and people will be injured or killed as a result. It is also inevitable that people will be injured or killed as a result of people driving cars. But in the long run, it's virtually guaranteed that self-driving technology will result in safer cars and fewer driving deaths. Computers can respond quicker, communicate well with other computers (hopefully) and they don't get drunk.

It's a fact of life that while accidents from ordinary drivers go unnoticed (ironically because they are so frequent), car crashes from self-driving cars will get covered extensively by the media, at least for the short term.

All this means that it will be a tough regulatory environment for self-driving cars in the US. In China, the government has more power to create regulatory policies around self-driving cars that take a long-term view without having to worry as much about lawsuits and other concerns.

Perhaps the fear of seeing the Chinese companies "getting there first" with self-driving cars will be an impetus for US regulators to create policies that are encourage the development of self-driving cars.

The issue is that self driving cars are naturally _worse_ than human drivers, and the only way to ensure that only the _good_ self driving cars get on the road is regulation.

Like flying, if anyone could build an airplane and fly it, flying would be more dangerous then driving. It's the FAA which keeps these things in check.

Now, if the various state DMVs could ensure that only the good human drivers get on the roads...

Actually, in the US anyone is allowed to build small aircraft and fly them completely unlicensed. (provided it's not at night or over a populated area)

I want the DMV to only license drivers when they are safe, alert, and driving defensively, so I want the DMV to only license autonomous systems when they can meet or exceed the bar of safe, alert, and defensive drivers. We can't enforce that humans drive that way after they're licensed, but we can enforce it with autonomous systems. We are already happy regulating driving, and we shouldn't lower the licensing bar to how people actually drive on average, "You get your license as long you drive better than the average Joe looking at google maps while he tailgates."

> We can't enforce that humans drive that way after they're licensed

Wait, isn't that what traffic police are for? Sometimes I wish speed limits were universally removed so that police will be forced to enforce the other 99.99% of traffic laws more strictly.

> Wait, isn't that what traffic police are for?

They just punish after the fact. I'm still able to make a poor driving choice that results in the injury or death of a bunch of people.

They also incentivize you to get your shit together before terrible things happen. After all, that's 50% of what punishments are for.

Sure, but they still don't physically stop you from doing something bad. It's still very possible, and actually easy to do it.

Maybe the word "can't" was a poor choice. My point remains that people pretty much always drive in ways that would fail a DMV driving test. Plus, I know I sure drive differently when cops are around too.

Government is way ahead of you there. The California DMV has been licensing self-driving cars for on-highway test for several years now. 19 companies have signed up. The National Highway Transportation Safety Administration had their meetings on regulations for self-driving cars last year and this year, and the proposed regulations are out. Basic rule: the manufacturer is responsible for all accidents of self-driving cars. Google and the big car companies are on board with this. Tesla has grudgingly accepted it.

> for US regulators to create policies that are encourage the development of self-driving cars.

How much more friendly of an environment do they need? The NHTSA has guidelines, and wants answers, but it hasn't shown a lot of prior restraint in this space.

These companies are slow to move because they know full well that if they kill enough drivers, their stock will fall so much that they'll be de-listed from any legitimate securities market in a matter of days.

This is a complicated space with a lot of new liabilities and civil responsibilities to consider, it _should_ move slow.

In Asia, to say it gently, people follow less rules... Baidu is going to have a hard time, as people take advantage when you are driving carefully.

Beijing could just mandate that the ring roads from 6AM to 10PM are self driving only, like they already ban trucks during that time. The cars will also have cameras galore, and who knows...if they are owned by the state (or a taxi company), they might as well just report bad human drivers to ensure good behavior.

Yea, I'm very curious how this will work out. Traffic there is madness and people use all parts of the road to get to where they're going. Maybe it will be simpler. Don't hit anything in front of you, don't let anyone merge, use any piece of visible road.

For those who are curious, two guys just did a motorcycle race through rush hour traffic in China [1]. It gives a good idea what the road conditions are like.

1. https://www.youtube.com/watch?v=axZC4inhu9c

How come no one has figured out that these self driving cars can be tested with light soft and fuzzy tops which only look like real cars, so if they crash into something, even at high speed, they wouldn't hurt anyone and couldn't be sued for a lot? They could then skip to testing much more daring maneuvers on higher speed than they would otherwise.

Unless I'm missing something the car needs to have an engine, that's heavy enough to kill people, even at moderate speeds, no?

Not if you go all-electric. Motors are light, batteries are heavy, but battery packs give you much more packaging flexibility than an ICE engine.

To an extent, but at those current draws, you do need integrated cooling in your battery packs.

But the engine can be surrounded by a lot of SOFT components which cause deceleration. What happens is that the outer layer will give way and the other cars will not be harmed in this deceleration. ESPECIALLY since I imagine most of the collisions will actually be glancing or scraping ones. This is strictly better than a real hard outer shell as now.

I guess the reason they haven't done it is that humans are required to be inside the car for now.

That's a... great idea. Self driving cars without passengers, where the hard components are close to the ground, and with soft padding everywhere else.

People who worry about the risk of accidents should also take into account the risk of falling behind technologically and economically. I'd argue that the latter would cause (or fail to prevent) far more deaths. If China wants to lead the world on this then more power to them. Maybe the US will step up their game in response.

People worry about accidents as if we aren't already causing thousands of accidents every single day. SDCs are inevitably going to reduce that danger.

Here in the US, self-driving cars will cause some challenges on the legal front.

In China, lacking the rule of law, it'll just be a calamity.

Don't forget that they hired the guy behind Google's self driving car, Andrew Ng: http://venturebeat.com/2014/07/30/andrew-ng-baidu/

Does it takes just a hire to compete against all Google's patents?

"compete" would imply Chinese recognition of US patents, which seems hit-or-miss.

Aren't you thinking of trade secrets by that statement? Patents would let anyone do it you just have to pay google their fee.

Andrew wasn't their self driving car leaders, he was their DL and ML leader, he didn't lead the SDC division specifically.

Andrew Ng wasn't really the brain child of Google's car. That honor goes to Sebastian Thrun.

Reply on news.ycombinator.com