Hacker News new | past | comments | ask | show | jobs | submit login
Uber self-driving trucks are now moving cargo for Uber Freight customers (techcrunch.com)
365 points by wil_wheat_on on March 6, 2018 | hide | past | favorite | 266 comments



For those concerned with jobs to be eliminated, it might not be as big a problem as we think for the trucking industry.

"Over the next 10 years, he said, we need to attract almost 900,000 new people to the industry.

Demographics is a big part of the problem. ATA’s research arm, the American Transportation Research Institute, recently updated its demographic data on drivers and found some 57% of drivers are 45 or older. Only 4.4% are 20-24 years old, noted Rebecca Brewster, president and COO of ATRI."

http://www.truckinginfo.com/channel/drivers/news/story/2017/...

In addition, there will still be much need for humans to take care of the maintenance of the trucks and manipulation of the cargo over the next couple decades. So there will likely be a significant number of truckers still employed, perhaps with increased productivity.

Also, certain other jobs will create large increases in demand for labor over the upcoming decades, in particular home care for the elderly and, if we get politics right, better & more individualized education for the young.

(A reminder) Lump of labor fallacy: https://en.wikipedia.org/wiki/Lump_of_labour_fallacy

---

Added in response to some comments below:

Yes, there will be challenges particularly regarding the distribution of benefits from increased productivity. The share of wealth garnered to capital holders will likely increase. People in labor force who have the skills to develop and maintain automation systems will certainly benefit. (That is why education and (re)training are crucial.) The rest may enjoy some, but probably not most, of the larger pie created by advanced technology.

However, the economic pie will be enlarged and there will be more to share. How the fruits of technology will be distributed will depend to a large extent on negotiations and politics.


Is the issue a lack of jobs or a lack of jobs that pay enough to prosper in our society?

BLS has the mean annual income for a truck driver at $43,000 (https://www.bls.gov/oes/current/oes533032.htm) versus ~ $22,000 for a personal care assistant (https://www.bls.gov/ooh/healthcare/home-health-aides-and-per...).

That's a huge difference, especially in a country with the health care system we have.


I think the main problem with the personal care assistant isn’t so much the pay is so low but rather it’s hard for people to actually pay someone full time (or even part time) to care for someone.


A full-time care taker is also often living in the house. So, even if they are are only getting paid $22k, that money isn't really going to any expenses.

They are likely getting paid more than that anyways.


I'm not sure how "often" but some do.

There might be people with special arrangements but if you are like the people I interact with then you probably engage with an assisted living agency for anywhere from $18-$30/hr. Let's take the low number as an example. $18 * 24 * 365 = $157.7K per year. How many people can afford $150k just for care? Not very many.

And $18/hr is on the very low end of the spectrum. If you're at $30+/hr that's $250K+ on up not including medical, insurance, etc. IME, caretakers are normally paid out-of-pocket, too. It's expensive business, especially if you have to do it for any period of time.


Assuming the actual work to care for an elderly with non-full-time need is up to 8 hours a day, paid at $12 per hour, the cost would be 8 * 365 * $12 = $35,040 per year. Other family members can hopefully cover the rest of hours if needed.

My question: What are the benefits of going through an agency when the cost could be over 100% higher than hiring someone directly? Hiring direct allows you to screen in person and makes sure there is a good fit of personality and other factors too.


My gut response is, this could be ripe for Uber-like disruption. I had relatives who used home-care assistants and another who worked as one, and the discrepancy between cost and compensation seemed quite stark indeed.


It has been tried.

That was the whole point of HomeHero[0] which tried to do that model but was thwarted by legislation.

Again, I'm no expert but I've been spending a lot of time in this area lately and ultimately while it's nice to think about it's also tough:

1. Scheduling. Finding someone to come by once or twice a week to do some chores and housework is easy. But if someone needs 24x5, 24x7, or some weird incantation of it then only a single person scheduled just won't work.

2. Turnover. I'm not sure how it is in other places but turnover here is high. Sometimes you might not even get notice when someone quits. That's why a lot of people go with agencies—they take care of all of that hassle as well reporting. Also, when you have people with memory problems, there are special people needed to take care of them. People who don't mind being cursed at or belittled. Some people just can't handle it.

3. Legal issues. IIRC, California does stuff where if you work more than 8 hours in a day you can make overtime—even if that's the only day you work. Some of the laws and handling as well as reporting on this stuff make (especially if you do have some sort of insurance or coverage for long-term care)... well let's just say I'm glad I don't have to do any of that.

[0]: https://medium.com/@kaleazy/theres-no-magic-in-venture-backe...


because you're already tired and taxed from dealing with a senior with dementia or a child with severe mental retardation, and there aren't enough hours in the day as it is.


if 22000 is the mean they are equally like to be getting paid less.

(well actually if it's the median it's equally likely but for practical purposes...)


Those numbers make sense at a glance. A truck driver, despite the popular image, more closely resembles specialized work than un-specialized. There are specific skill and documentation requirements to become a Truck driver. The healthcare work you linked to just requires a high-school diploma and no certs, if it's a job like caring for the elderly then the main factor driving up the wage is a shortage of people willing to deal with the challenges of working with an elderly person which can include rude or mentally ill families/customers, excrement, and being comfortable cleaning and/or bathing another human being.


Trucking is absolutely skilled work. It's crazy that people think it isn't.


You start by saying how a personal care assistant is rather unspecialized work and then go on to ruminate on the peculiarities of the job that most people wouldn't want to deal with.

Why isn't trucking an unskilled job just with negative features (long hours alone, danger to health given the erratic hours) too?


>You start by saying how a personal care assistant is rather unspecialized work and then go on to ruminate on the peculiarities of the job that most people wouldn't want to deal with.

Correct. Being willing to endure undesirable work isn't usually included in the definition of skilled labor unless that tenacity is accompanied by another requisite skill or certification. I'm not trying to argue that it isn't a socially valuable skill, just that given the bar for entry it isn't surprising the wage is that low.

>Why isn't trucking an unskilled job just with negative features (long hours alone, danger to health given the erratic hours) too?

Because you've skipped over the training and licensing required to legally operate as a truck driver. You need a Commercial Drivers License if nothing else and many states require the completion of specialized courses before awarding an applicant their CDL. That CDL represents a specialized skill-set without which a person simply won't be permitted to operate as a trucker. Physical endurance and mental tenacity are critical traits in this instance too but they're far from being enough to get you hired.

Edit: Something else I glossed over, but which seems relevant because this is predominantly a tech oriented forum is that the average Cost of Living in America is in line with the wage being paid to the Home Health Aide. It might not be enough for San Francisco or a coastal area, but it's enough in general to pay rent and save some money on the side for additional schooling or whatever purpose appeals.


Generally, the idea of stifling innovation to "preserve jobs" is harmful to society. Upon choosing a career, you should consider the longevity of that career. Some people will still pick wrong for a multitude of reasons. Hopefully, those people can pivot. It is unreasonable to hold the entire population back to continue employing those in a specific field. If we consistently did that, we'd be in the stone age and our quality of life would be far poorer. This will help future generations drastically even if it hurts us a bit.


> Upon choosing a career, you should consider the longevity of that career.

Isn't it lucky we humans are so very good at predicting the future!?

When I was taking computer vision courses little more than a decade ago, self driving cars were pure science fiction. Part of me hopes software development is the next job we automate away, just so everyone in forums like this who are so impressed with their own excellent life choices can see what it's like to be on the other end of this.


>Part of me hopes software development is the next job we automate away

We are hard at work automating software development. Thanks to massive advances in programming language design, IDEs, the invention of module systems in combination with the massive success of OSS in that space and the proliferation of massive frameworks, something that took a year of work two decades ago is now at best a weekend project of a much less skilled person.

The thing that keeps us all employed is the incredibly elastic demand. And of course that the work to find and describe requirements is unlikely to be automated away anytime soon (and already often takes more time than actual implementation)


> Thanks to massive advances in programming language design

What massive advances in programming language design have we had since 1998!?


While a lot of the real advances in language design are not new, popular adoption of stuff like functional programming has happened in large part in the last twenty years. E.g. list comprehensions in Python were not introduced until Python 2.0 (2000, which just squeezes into your range). And Java is starting to get similar features, which is great for companies who have invested in the language.


> Some people will still pick wrong for a multitude of reasons. Hopefully, those people can pivot.

We are lucky as software engineers, but I was also making the point that if your line of work gets automated, hopefully you can pivot. This applies to everyone, and is similar to when you are no longer enjoying your field of work. Changing careers is not always an option, but let's not forget that it usually is when we talk about automating an entire industry and putting them out of work.


So give them the support they need to retrain. We're creating so much value with these automated systems - who is the beneficiary? Surely the compassionate thing would be to use some of this resource to help those that have been trodden on by progress/


I'm hoping for a future where the machines do all the work required to provide everything human beings need to survive, AND where we've figured out a reasonable way to distribute resources in a world where not everyone will be able to "contribute to society" by spending huge amounts of their waking time doing something they'd not otherwise do, in exchange for tokens that can be redeemed for food and shelter.


Yes, but that would be communism, so let's just let people who were born into or stumbled into their wealth amass more of it. I think in general we need to start being more thoughtful about how the things we do can sometimes be hurtful to billionaires.


Well, while you’re joking the original idea of communism and what I’ve been taught about Lenin was about excatly this - communism is inevitable because of technical progress. Call it what you want - universal income, handouts, we will end up in communism.


And yet all implementations of communism has failed so does this mean technical progress eventually leads to societal failure? Also I believe we will end up with socialism, not communism.


> And yet all implementations of communism has failed so does this mean technical progress eventually leads to societal failure?

Only if you believe that once an attempt at something has failed, all future attempts will also fail.

> Also I believe we will end up with socialism, not communism.

Socialism is worker control of the means of production, and is an intermediate step along the route to communism, a fully classless and stateless society.


So if your line of work gets automated, you retroactively "picked wrong"?


It's not far from it. Software development is in small company with respect to the need for constant re-education and skill maintenance. The vast majority of developers with skills from just 10 years ago are wholly unemployable right now.


You mean somebody who knows how to develop in Java, PHP, Python, Perl, C, C++, Javascript, Ruby, LISP - all languages developed more than 10 years ago - is "wholly unemployable right now"? Doesn't seem to match my experience even a little bit. Programming practices also hasn't changed that much. In fact, there are still jobs for COBOL and Fortran programmers - not that many, but there aren't that many COBOL and Fortran programmers either nowdays.


There's a shortage of COBOL developers that are needed to maintain and update legacy financial mainframes that still don't make economic sense to replace since they operate at critical infrastructure points 24/7 and over the years have become pretty bulletproof.

Replacing these systems is extremely difficult since they're so relied on a replacement needs to be nearly flawless form the start. If you stumble upon on a bug in your international wire transfer code you could be messing up a $10 billion international aid wire that was meant for hundreds of factories.


Um, no.


Despite being a coder, I almost desire tech work to become replaced by machines at some point, with the only options of retraining to be prohibitively hard and leaving us in the dreadful position of these workers, just so I can watch people with attitudes like this get it handed to them.

The attitudes we have towards the unskilled, low-skilled, or those who are being replaced and made redundant by the systems we develop is sickening.

Describing someone spending extensive amounts of time and money retraining and having to abandon a career they've done for decades as "pivoting". Will you provide for their family in the years it takes them to be proficient at a skill that they can use to replace their income and support themselves? Or are we just all snidely sitting here with our tech know-how preaching to those who grew up in a different time, or with different opportunities, or who couldn't have ever predicted the scale to which their livelihoods would become obsolete?

"Hurts us a bit". Don't even speak like you're in the same boat as these people.

Don't get me wrong, progress will happen, but we need to make it socially aware and compassionate progress, not tough luck you didn't pick a the right career and now your life is fucked progress.


I don’t really understand you mate, what are you suggesting, stop technical progress? Sure it harsh and one day we will be next but so be it, it’s life.


I'm not sure where your confusion is arising from. antihero included a preferred alternative:

> Don't get me wrong, progress will happen, but we need to make it socially aware and compassionate progress, not tough luck you didn't pick a the right career and now your life is fucked progress.

That alternative is not stopping technical progress. It's providing for those whose jobs have been made obsolete by that progress, so that they can rebuild themselves. And keep in mind that not doing so carries the cost of having members of society who are unable to provide value beyond their physical existence.


Did you read my comment? Progress is awesome and inevitable, but how we deal with it and how we deal with the people who's careers are made redundant by it is what we need to discuss. We need to give people options, whether that's retraining and financial support, universal basic income, or whatever. Currently the money just goes upwards - the progress, innovation are owned by the elite as intellectual property. We need to find a different way where the fruits of this progress are distributed to benefit humanity as a whole, not used to further exploit the masses.


Although I agree with you, the problem (in the US at least) is that pivoting is nearly impossible for a large group of people. Our educational system is expensive and generally has terrible quality, and we don't have very good social programs to help float people while they are getting re-educated.

If we push innovation without changing these problems, it's not gonna be pretty.


If you don't need shiny credentials, educational system is much less expensive - there are community colleges, courses, free or low-cost online education, etc. Most costly educational requirements arise when coupled with licensing requirements - if you need an official paper, you're gonna pay.


When I graduted high school during the .com bust my father told me not to major in CS in college because there wouldn’t be any jobs for me when I graduated.


Absolutely. Automated trucks will allow shipping costs to be much cheaper which will stimulate the economy in 1000 different ways


Shipping is already cheap. Especially bulk shipping.


transportation costs add to the cost of literally everything


Of course, but the point is that they don't actually add all that much cost. It's a few percent. A nice gain if it goes away, but not something that is going to fundamentally change the economy.


A few percentage points on the entire economy? Not good enough?


What percentage of a trucking company’s costs do you think are paid to the driver?

It won’t lower shipping costs that much.


There's plenty of surveys on this[1]. In this one driver wages and benefits make up around 40% of total cost per mile. That's why self-driving trucks are so lucrative.

[1] PDF warning, see page 23 http://atri-online.org/wp-content/uploads/2016/10/ATRI-Opera...


This is a great resource, but there are a few things wrong:

1. Fuel prices were low in 2015 and have been rising since then, so 39% of marginal costs is likely a peak.

2. The chart you referenced is only for marginal costs, and doesn't include fixed costs associated with a trucking company. The actual percentage of driver pay to the overall operating cost of a trucking company will be lower once fixed costs are factored in.

3. Automated truck companies will also want to capture a fraction of that 30-40% that is paid to drivers.

So in the end, shipping costs will likely drop in the 10-20% range, which is substantial, but it isn't enough to be revolutionary.

My point wasn't to the value created by having automated trucks, but to counter the point that shipping costs would drop significantly.

Also, since shipping costs are a small percentage of the price of most consumer goods, everyday purchases will not be affected much either.


"Stimulate the economy" and "provide prosperity" correlate only very weakly, if at all.


"Stimulate the economy" and "provide prosperity to some billionaires" correlates very well. :-/


[flagged]


If you want to avoid downvotes from “the hivemind”, you'll have to post more substantively.

https://news.ycombinator.com/newsguidelines.html


If you could actually predict and pick with more than random success, you'd make more money predicting and picking then you would working in whatever you predicted.


Yea, every industry makes this claim and it either means that no one wants to do the job because it is harmful or doesn't pay well enough. I don't think it is a good counter argument to the automation of one of the largest industries that employ people in the United States. Even if you take the claim that it won't impact jobs you have to look at other people in the ecosystem and how their jobs will be impacted (automated trucks don't need to eat or sleep and that means those truck stops will stop existing).


The rate of change matters far more than the amount of it.

https://ourworldindata.org/employment-in-agriculture


> "Over the next 10 years, he said, we need to attract almost 900,000 new people to the industry.

The pay of truckers has, in inflation adjusted terms, shrunk by more than half since the 80s. They mostly need to bring in more people because their turnover is ridiculous.


Automated Teller Machines (ATMs) were set to put Bank Tellers out of a job, but since its introduction, the number of Bank Tellers increased [0]. While I can't guarantee that something similar will happen for truckers, automation does not guarantee that fewer humans will be employed doing that specific job.

[0] http://www.aei.org/publication/what-atms-bank-tellers-rise-r...


I'd love to see that graph continued for the last few years.

With online banking becoming more mainstream and online-only banks picking up steam, I've personally seen many local bank branches close up, and have a hard time believing that.

> So the transition–what the ATM machine did was effectively change the job of the bank teller into one where they are more of a marketing person.

Not sure if such an alternative position can be found in a cutthroat B2B market.


Bank Tellers have however as a result turned into entry-level-retail with often impossible quotas for sales.


I don't have data but my gut feeling says that the numbers of Bank Tellers has peaked. In my country (Netherlands) a lot of small branches of banks have closed. The general trend is towards self-service / online.

Anecdotically, I went to the bank across the street 2 years ago to ask some question about insurance. I was told to go to the website and look up the information there. The branch is now closed.


Most grocery stores don't even have self-check stations. It's clear that these open more opportunities for loss than they generate returns in increased shopper throughout and return visits.

Meanwhile, my mother still curses the personal computer revolution for taking her job as a transcriptionist. Everyone types today, so specialists are rarely needed.


This isn't actually correct, or at least it's not a correct interpretation of the data. The absolute number increased, but it decreased as a percentage of population, which is the relevant number here.


>much need for humans to take care of the maintenance of the trucks

Is that really true? It seems electric trucks[0] quickly follow from autonomous trucks and from what I understand, electric vehicles are mechanically much simpler. That leaves truck operators with not much need for a greasemonkey wrenching on a transmission when there is a) little/no grease[1] and b) no transmission[2].

Tire & body shops are about the only maintenance jobs I can think of that won't be reduced/eliminated by the coming changes.

[0]https://www.tesla.com/semi [1]https://www.mckinseyenergyinsights.com/insights/impact-of-el... [2]https://www.roadandtrack.com/new-cars/car-technology/a120190...


There’s also a lot of drivetrain and suspension related maintenance, but it’s less than engine maintenance.


Slight tangent, how's the Tesla semi-truck been doing? When I was looking over the specs the range seemed low compared to a regular semi but I wasn't clear on if that was acceptable because they're targeting a particular subset of the semi-truck market.


The range is fine for single-driver operations, as the driver still has to stop and take rest breaks equal to or longer than the stated time to recharge on a 'megacharger'.

I suspect that there will be deals done (if not already in progress) between Tesla and/or Fleet companies with truck stops across the country to install (paid) megacharger stalls at appropriate distances.

Even if charging time is slightly longer, it might still work out equal if the power of the truck lives up to expectations, and the truck no longer needs to slow down when going up hills.


It seems like a strange time to buy a nationwide chain of truck stops[0] with the oncoming electrification of vehicles. Unless... there's an opportunity to provide Supercharger access every few hundred miles.

[0]https://www.bloomberg.com/news/articles/2017-10-03/buffett-s...


It's not a major point, but I'd be curious to know the average age that the over-45 drivers started driving. Did they drive from their 20s, or did they pick up truck driving as a 2nd or 3rd career?


> Did they drive from their 20s, or did they pick up truck driving as a 2nd or 3rd career?

Just one data-entry point (and it's from Eastern Europe, not the States): my brother picked up trucking in his mid-30s, that is 5-6 years ago. Until then he had been a farmer, but as the agricultural subsidies in my country can't rival with those given to French farmers (for example) there was no way for him to earn his living just by raising cows (even though he loved the farming work). Until this trucking job he had never been out of the country, in the meantime he has seen almost all of Europe up to Western Turkey. It's not a job for everybody, though: a month or so ago he was robbed of 240 liters of diesel fuel while he was in his sleep-pause somewhere in Bulgaria, he had just been waiting for 20 hours a short time before that at the Turkish-Bulgarian custom point, and he's usually gone 2 or even 3 weeks at a time.


Annecdotal evidence but I know two people who started in their 50s. It’s a good second career for people who are getting too slow for manual labor.


Can the history teach us anything here? What had happened when the technology called "China" became available? How did it affect jobs elimination?


> However, the economic pie will be enlarged and there will be more to share. How the fruits of technology will be distributed will depend to a large extent on negotiations and politics.

I think this is you ceding your original comment. I think most people clearly understand this is merely a (potential) distribution problem.


> In addition, there will still be much need for humans to take care of the maintenance of the trucks and manipulation of the cargo over the next couple decades. So there will likely be a significant number of truckers still employed, perhaps with increased productivity.

There’s already an industry that performs these tasks so there’d be a small percentage of absorption of current truck drivers.

You also assume there’d be more trucks, but there’d likely be fewer, at least at first. Since a fully automated truck could operate substantially more hours than with a driver. Drivers have hours of service that restrict how much they drive per day and per week.

Also, on a macro scale automated trucks wouldn’t be adopted if they didn’t displace workers.


>some 57% of drivers are 45 or older. Only 4.4% are 20-24 years old

For reference:

Assuming a driver starts working at age 18 and retires at age 65, and given equal distribution of workers over all ages, you would expect 42% of workers to be 45 or older, and 10% of workers to be 20-24 years old (if you assume 21-70, then 51% should be 45 or older).

Just in case anybody else was wondering what those numbers mean. As an aside, this projection from the same presenation looks slightly suspicious: http://www.truckinginfo.com/fc_images/news/l-driverturnover-...


Did you factor in truckers’ short life expectancy?


anecdotal, but each summer I spend time at a hotel in the Austintown/Youngstown area which is a drop off point for newly "graduated" students from a trucking school.

the old timers are there too and some are there to pick up their partner as many long distance services have two drivers per cab. the old timers are a great source of information and if anything stands out its how many new drivers don't cut it. most leave on their own but others are pushed out for not being good drivers. good means not only a safe driver but properly courteous to everyone you meet.

quite a few get into it on the storied romance aspect of it, those stories that convey that there is something special in it, one with the road, etc. others because they don't understand the complexity of the job. it is not just driving.

so I am really curious how much work moves to dock yards and better yet, what processes are being put in place to handle all the interaction with local enforcement agencies because truth be told there are parts of this country where the locals stop trucks fishing for something to ticket and generate revenue on. how will self driving, no one aboard, eventually be handled?


The book The Long Haul: A Trucker's Tales of Life on the Road captures a lot of stories with feelings like the ones you described; romanticism that stems from on identifying heavily through their work. Here is one of those stories from the book [1]; I'd recommend people in this thread give it a read, it is worth knowing the lives of so many Americans before it is automated and stories like these disappear.

https://longreads.com/2017/09/21/a-high-end-mover-dishes-on-...


I rarely see justification for the idea that lump of labor is actually fallacious. Historically big die-offs have led to higher wages for survivors.


The counter-example would be the Industrial Revolution. Compared to 150 years ago, nearly every job has had the amount of work necessary to do it dramatically decreased, yet wages increased dramatically.


If those high wage earning survivors are also paying high wages to all the other survivors, then are they actually better off? My gut reaction is no, less people means less work is being done and you lose a lot of economies of scale.


Few people pay wages, many people collect them.

But I don't think history will be useful, because depopulation also causes on increase in agricultural productivity (per person) that is a huge confounding variable.


Few people pay wages directly, everyone pays wages indirectly.


Let me answer your question with another question: looking at the world as it exists today, would you rather be a person of average means in a high-wage country, or a person of average means in a low-wage country?


Are all the products I purchase produced within the high-wage country, or are many of them produced in the low-wage country? If so then I'm shielded from paying high-wages to a great degree and so get the best of both worlds.

Lots of people retire to (or work remotely from) low-wage countries (or regions) because their money will go further there.

Manufacturing keeps moving to the next poorer country as wages rise because people don't want to pay more for their t-shirts or other manufactured goods as countries industrialize.


That only goes so far. You can't offshore building your home, for instance.


> However, the economic pie will be enlarged and there will be more to share. How the fruits of technology will be distributed will depend to a large extent on negotiations and politics.

If recent history is any indication of how this will turn out, it's not going to be pretty. The rich get richer and everyone else gets screwed.


> The rich get richer and everyone else gets screwed.

Not actually true. a more accurate telling is: the rich get richer and the poor get richer too. You may commenting on the fact that the rich are getting richer FASTER than the poor are getting richer, and that's an interesting phenomenon to comment on, but it is undeniable that everyone is getting richer, almost regardless how you slice the data.

Worldwide, extreme poverty is on an accelerating decrease, and is set to be completely eradicated by 2030. In the US and in Western economies, poverty, which used to be common only 80 years ago, is almost completely erased. Meanwhile those who have low income, but are above "the poverty line", continue to see growth in wages, in standard of living, and in lifespan and in quality of life. Objectively.

Why are we are inundated with the false idea that "things are so bad" by mainstream media? That narrative sells papers and clicks. But it's not supported by the data.

Whether we all would benefit from rebalancing the wealth distribution is a different question.

reference: Humanprogress.org, and Steven Pinker's new book, "Enlightenment Now" (highly recommended)


I have always wondered how we can reconcile that things "are" objectively so great and yet "seem" subjectively so bad. In my experience both online and off, there is a pessimism about work, poverty, and basic security that persists (or is even getting worse) despite these changes. To my eyes, it has to be larger than just the state of the media.

We have a few of things we can quantify, and that are often brought out in discussions such as this one. Healthcare outcomes, wages, life expectancy, basic material goods, access to education, casualties from war, etc.

I heard another commenter here talk about the human experience being understood as a vector, with twenty or thirty dimensions. Most of those are moving in their positive directions. But the problem is that when God made the human experience, he crafted it with uncountably many components, most of them themselves unquantifiable.

Unfortunately for us "objectively" exists only in that limited set, not in the greater whole. "The number of species going extinct per unit time is more than it has ever been, maybe ever." What is the cost, paid in pessimism and hopelessness rather than dollars, of knowing that? Does it counteract a 2 month increase in my projected lifespan?


> how we can reconcile that things "are" objectively so great and yet "seem" subjectively so bad

Because one of the things that has been fueling GDP growth is disintermediating community and family support systems. Health care spending is up because you don’t have a family home where the older generation slowly ends up and where the younger generations can share elder care burdens. You don’t have people cooking for the extended family. You don’t have neighborhood sharing of homemaking tasks. Etc.

These things have been disintermediated in favor of nuclear family and individual solutions provided by commercial providers. Because that leads to GDP growth and GDP growth drives culture.

But even though you might be getting a slightly bigger piece of a much bigger pie, and even though fewer people might be starving, a much higher percentage of your personal self care burden is falling 100% on your shoulders. And that means you have more pressure on your earning potential, which means you are more sensitive to changes in employment.

Essentially, a larger portion of your well-being is bottlenecked through your checking account. So even though you may be “more taken care of”, the subjective (and possibly objective) precariousness of your well-being is much higher.


> Why are we are inundated with the false idea that "things are so bad" by mainstream media?

Maybe because it's not false and it's actually happening to an alarming degree in the U.S.?

https://hbr.org/2017/10/why-wages-arent-growing-in-america

http://www.pewresearch.org/fact-tank/2014/10/09/for-most-wor...

https://www.bloomberg.com/news/articles/2018-03-01/america-s...


Re training will fail. This is the major fatal flaw of the current economic system.

Supporting data: The writing has been on the wall in glowing neon ever since the advent of MOOCs.

MOOCS have many near magical attributes that we have long sought in education

1) The students regularly self select into the courses they want, as opposed to schools where you don't have a choice.

2) Courses are free

3) Courses are high quality, and improving with testing

4) Courses are media rich, and use multiple types of stimulation for stronger impact when studying

5) Honorary mention - gamification

6) Material is always available, at any time - you are not limited to teachers being available

7) You even have online communities you can reach out to, to discuss your work

And yet they have about a 10% completion rate.

Let me underline - thats a 10% rate with a self selected pool of interested applicants.

MOOCs are literally magic.

IF you are talking about re-training then you are missing the Big Picture: The re-training and skill up revolution has come and gone.

We should currently be seeing a large rise in the number of people who are self actualizing, and learning skills in a wider range of industries.

-----------

In economic terms, the failure rate of MOOCs indicate that retraining attempts will result in very poor retrenchment of workers.

Most jobs lost will result in under-employment, increasing the concentration of capital in the hands of capital holders.

This will have a knock on effect on the America old style university system, as this will become the only path to a higher paying career over poor paying service sector work.

----------

The retraining model which will work will require people to be allowed to train non stop on material, and the ability to focus and work on a subject till they have gained the necessary mastery.

I doubt this will be possible, because a family earner cannot take weeks off to master a skill only to have to hunt for jobs in a market where younger students already abound.

Consider how many people want to get healthier, apply to gyms - and then stop going.

Re-phrase the retraining problem as an attempt to get all of humanity (or a slice of it) fit.

Also since the tools of technology allow for hitherto impossible levels of orchestration and conversation management, the barrier to outrage driven organization and outrage has been set much higher, which will lead to less negotiation power during the eventual distribution of wealth


The funny thing in this US-centric view is that we have such a cargo since a century where trains play the role of "self-driving cargo".

In europe, the decline of cargo on tracks (trains) is also partially owed to subsidiation of automobile traffic and highway expansions. It is a pity, because from the raw numbers trains are much less expensive then tracks.


I think that's misplaced, freight is the one area where NA in general and US in particular absolutely blows every other region out the water. Circa 2010 Rail was ~40% of US freight by ton-mile, whereas for the EU it was ~17%, and the US had 10x more tonne-kilometer freight rail than the EU (beyond China and Russia, with India a very distant third) (nb: China carried more tonnes total but shorter distances)

Of course a major reason for that is that the US have (literally) tons of commodities, a significant plurality of the tonnage is coal. I expect roughly the same of Russia, China, Australia and India (the other members of the tonnage top 5, Australia notably haults extremely heavy stuff for very short distances as it's about on par with the EU for tkm but waaay ahead in total tonnage).

Hell one of the reasons passenger trains is shit in the US is that passenger rail does not get priority over freight and thus you get the country's lone "high-speed train" giving way to slow freight leading to an average speed of 130km/h, slower than some non-high-speed intercity service in Europe.


> Hell one of the reasons passenger trains is shit in the US is that passenger rail does not get priority over freight

They do have priority:

> (c) Preference over freight transportation. -- Except in an emergency, intercity and commuter rail passenger transportation provided by or for Amtrak has preference over freight transportation in using a rail line, junction, or crossing...

http://codes.findlaw.com/us/title-49-transportation/49-usc-s...


Passenger rail is supposed to have priority by law, but in practice they don't always get it. Freight carriers don't always yield, enough so that there are ongoing legal battles. You can check Amtrak's historical On-Time Performance and causes of delay here: https://www.amtrak.com/about-amtrak/on-time-performance/

My quick sampling of ~10 trains came up with an average around 65% on-time during the last year. The only one I've ridden, the California Zephyr, is less than 50%.

http://blog.amtrak.com/2015/02/message-amtrak-regarding-time...

http://trn.trains.com/news/news-wire/2017/03/24-judge-sides-...

http://www.jconline.com/story/money/2017/08/24/csx-digs-its-...


Yeah I've ridden the Zephyr from Chicago to Colorado (and back) a few times, and it was off-schedule about half the time, one of them was at least 4 hours of just sitting there in the train in the middle of nowhere.

One hour I can understand, four is ridiculous. The trip is pleasant and the views are great but for me to take it again they're going to have to do better than 50% on-time.


I used to live on a small town on the Empire Builder line a few hours west of Chicago. It was once a day each way, and we'd get one direction in the morning, and the opposite direction in the afternoon. Or that was the theory. Often the morning train was so late that when you heard them sound the horn you wouldn't know if it was the afternoon train or the "morning" train.


The "Late Shore Limited" I've taken has on-time performance of ~50% and train interference as 50% of delays. That sounds about right.


Apparently, according to your 2nd cite, that law was determined to be unconstitutional.


Sounds like something that could be enforced by giving the passenger trains the ability to lock out switches further ahead of them. Like, say, an hour ahead of them. (Though, without signalling the particular switches that lower at-grade crossing guards. It’d be a rather fraught thing to engineer, all-told.)

Sure, you couldn’t run densely-packed passenger trains very well on such a system... but they’re not densely-packed. If they were, the system would have been a success and could then be turned off. ;)


This is starting to sound like distributed systems ... how do you prevent deadlock? https://wiki.openttd.org/Deadlock :-)


What happens is that the freight trains gives Amtrak a window for its timetable. In that window, Amtrak gets priority. Miss the window, and Amtrak has to wait for freight trains. There's a separate issue that some freight operators will decide to screw Amtrak over (even within their nominal window) and just eat the fines instead.


In Tucson AZ it was a constant problem that freight trains would either blow through town at full speed or crawl through level crossings blocking traffic rather than obeying speed laws. It was always cheaper for them to pay the fines than to comply.


I read "priority" from the GP as "commercial/investment priority".


I'd be curious to know if freight travels shorter distances on average in Europe though.


The data mostly looks like population data then, if US, India, and China are the leaders. What would it look like for ton-mile per capita (if that's actually a statistic)?


> The data mostly looks like population data then

The US has a smaller population than the EU (~300m to ~500m). So does Russia (150m). Australia has a significantly smaller population than the EU (15^H^H25m as golemiprague noted).

It's not population data it's mineral commodities.

> What would it look like for ton-mile per capita (if that's actually a statistic)?

The US would climb to ~15x the tonnes-kilometers per capita of the EU, and australia would shoot up to about the same ratio over the EU (owing to close tonne-kilometer but 1/20th the population). Russia would shoot up as well. Only China and India would get a lower rank overall, and not enough that the EU would come close to edging them for China at least (a bit under 3x the population, but a bit under 10x the tonner-kilometer of rail freight).


Australia has close to 25m


Indeed typo'd in the first paragraph. Fixed. Note the ratio is correct in the second one.


Europe’s population is ~750m.


But the EU is not Europe.


The numbers I have are specifically for EU27, not for the entire continent, and EU27's population is about 500 million people.


The main issue has actually been prioritization. In Europe the rails are fast and well-maintained, which is great for passenger operation but expensive for very heavy freight operations. Passenger operations are also prioritized very heavily at the expense of freight operations on the rail network.

In comparison, the US rail network is slow and low quality, which makes freight operations extremely attractive but is absolute garbage for passenger operations. US rail freight competes well in its market niche, where it is faster than oceanic shipping but much cheaper than air freight.


I moved from western Europe to eastern Europe and public transport by train is absolutely horrible. I still hope that one day I can catch a train that arrives on schedule. We're not talking a couple of minutes delay, we're talking multiple hours. Even when the train finally departs, it'll pick up delay along the way due to poorly maintained tracks and bad scheduling. All of this turns 6 hour trips into 10 hour trips.

I am well aware that this is anecdotal. But from my point of view, not all rails are fast and well-maintained in Europe. Based on my personal experience, that statement applies to western Europe.


>> Based on my personal experience, that statement applies to western Europe.

"Western Europe" is almost always what Americans are referring to when they say "Europe," for various biased reasons.


not build for high speed operation != low quality.

High Speed Rail is just a different style of construction, etc - but quality is subjective.


American rails are decidedly low quality. [The accident record is much worse](https://www.nytimes.com/2015/05/21/world/europe/railroads-ar...), and the freight networks spent many years dismantling tracks and electrification to improve the bottom line.


HSR really depends on straight and level right of ways, along with some better signaling equipment and more (complete?) separation with road traffic. China builds a lot of its HSR tracks elevated with lots of tunnels to blast through hills and mountains. Standard track can stand to meander more for less construction costs.


Why is slow and low quality rail great for freight?


Low maintenance costs mean cheap transport. Coal doesn't care if the trip takes an extra day.


In the U.S., rail freight volume appears to be going up and to the right, measuring both by distance-weight or just by total number of containers moved any distance: https://www.railserve.com/stats_records/freight_railroad_tra...


You get advantages from trucks in terms of flexibility and lower inventory due to faster shipment times. It doesn't make sense for things like coal, sure, but there's very real advantages you gain through shipping by truck instead of rail.


It is a pity, because from the raw numbers trains are much less expensive then tracks.

Often, the flexibility and reduced latency of freight by truck is worth the increased per-mile cost.


Freight by truck is worth it because of flexibility and reduced latency, flexibility and reduced latency are worth it because of just-in-time production, just-in-time production is worth it because of less inventory, less inventory is worth it because of less money locked up in inventory, less money locked up in inventory is worth it because of better cash flow on the quarterly reports, better cash flow on the quarterly reports is worth it because it makes the stock market like your company more, the stock market liking your company more is worth it because ???

We might just find out the things we thought are worth it, are not really. It works the other way around, too, of course, which is that the increased per-mile cost is still way below what it would have to be with the externalities accounted for.


Low inventory is also worth it because it uncovers operational inefficiencies. So it not only helps you become more profitable, but also more reliable.


Doesn't it also mean that when things go wrong, they go really wrong?


> ...less inventory is worth it because of less money locked up in inventory...

Less inventory is worth it because space is money, especially on a retail shelf.


We might just find out the things we thought are worth it, are not really.

Then give a concrete quantification of that. It's a very complex question. Entire careers have been built on how to price something complex. That's called "economics."


See how this is an example of how the need to take a swipe at some group ends up making you look foolish?

Your point was that trains are "self-driving cargo". If you'd left it at that, this would have been a clever comment.

Instead, you look like someone who doesn't realize the US has massive, massive, massive amounts of train-hauled cargo because all he knows is we don't use commuter trains as much as other countries.


The video ad is so crazy -- we still need a human to lock/unlock the cargo?? and its good for the truckers because they can go home early to their family??

The first part can be trivially automated and the second is just stupid. Truckers get paid per hour of work, less work == less money.

AI-self-driving has arrived but to make the transition look "humane" as in the AD is crazy. Almost insults the intelligence of these guys who have been driving our cargo around for whole generations.


Truckers generally get paid per mile, not per hour. They are also heavily regulated as to the number of hours they can drive (why you see so many husband/wife driving teams).

One can easily foresee a future where a self-driving truck has a driver in it who is only driving 'regulated' hours for a small portion of a trip (say during high traffic, highway entry/exit, etc). The driver could spend the rest of his time sleeping, reading, or coding - he'd basically be a passenger.

If a truck can haul for 20-30 hours without stopping for overnights, that dramatically changes the economies of fast freight.


Coding is a very HN comment :)

More realistically, he can do the trucking office work from the truck.


Truck drivers don't do office work, really. All of the documents like bill of lading, etc are taken care of by the office workers of the company the driver works with.

They have to do paperwork like logging hours, doing a pre-trip inspection, etc. This is done during off-time while the driver isn't driving (for obvious reasons). Usually on the driver's seat with a clipboard on the steering wheel :D


they don't write code while they're driving right now either.

but if they suddenly started sitting in a truck doing nothing all day instead of driving, it's a lot more likely that some office duties would be shifted to the truckers rather than the truckers taking on totally unrelated projects.


This was my thought.


> One can easily foresee a future where a self-driving truck has a driver in it who is only driving 'regulated' hours for a small portion of a trip (say during high traffic, highway entry/exit, etc). The driver could spend the rest of his time sleeping, reading, or coding - he'd basically be a passenger.

That kinda already exists though, with team driving. Two pilots relaying gives ~22h of driving per 24h plus 2x30mn slack (under US regs, though there's still a limit of 70h/8 days followed by 34h rest per driver).


> That kinda already exists though, with team driving. Two pilots relaying gives ~22h of driving per 24h plus 2x30mn slack (under US regs, though there's still a limit of 70h/8 days followed by 34h rest per driver).

So at minimum labor costs could halve... Sounds like a huge productivity boost.


> So at minimum labor costs could halve... Sounds like a huge productivity boost.

It's not a change in productivity at all.

Also

> Truckers generally get paid per mile, not per hour.


One person getting the job of two people done is indeed a [giant] productivity boost. Compensation will surely change along with the industry, but there's not a reason to suspect drivers will be the winners in this.


You typically pay a surcharge for a team load, usually because you effectively move your load 1000 miles a day instead of 500. For certain types of loads, team drivers are how you make up for manufacturing delays.


rest of his time sleeping, reading, or coding

Problematic in at least two ways, probably three.


Even if (and that's a very big if) we get to a point where we decide we need someone behind the wheel of these truck, but it's okay for them to be staring at a laptop the entire time, I can't see a whole lot of quality code coming from someone who is sitting in the cabin of a truck with one eye constantly on the road.


That's why you hire 10x coders to do it, so that the worst you get is 2-3x.


Just when I thought there wasn’t anything worse than the open office plan.

It brings a whole new meaning to the word “shipping”.


> Just when I thought there wasn’t anything worse than the open office plan.

Eh. A truck cabin sounds better. You can have music without needing to wear cans, you don't hear jim's mastication or janice's burping you don't get assaulted by emily's lack of hygiene or randy's mix of aggressive cologne and smelly feet, you decide what temperature you're most comfortable at, you can work without pants, …

The main issue would be 'net connectivity.


I'm not kidding: I've imagined exactly that as a career change more than once. Live on the road, see new places every day, hack on open source projects, write, spend time with someone with similar interests (the hardest part). What's not to love?

Even removing the autonomous angle, I've thought about driving during the day and having laptop time for the same activities during rest periods. Probably seems weird, but I'd be in if I could make it work. I do my best thinking behind the wheel, and I've solved a number of engineering problems on road trips. It seems ideal.


> I've thought about driving during the day and having laptop time for the same activities during rest periods. Probably seems weird, but I'd be in if I could make it work. I do my best thinking behind the wheel, and I've solved a number of engineering problems on road trips. It seems ideal.

I hear trucking causes quite a bit of fatigue.


Why? I'm speaking of a completely autonomous truck, not some level 2 adaptive cruise control. The 'driver' could be sitting on a couch or at a desk in his sleeper cab the whole time.


I think you've missed this one - the typical long-haul truck driver drives somewhere far enough away that they can't return home and must stay overnight (sometimes sleeping in their trucks) before they can drive again.

This system seems to indicate that the automated trucks do the long hauls between cities and the human drivers do the shorter hauls within/around cities.

If that's the case, this is a good thing.


this is a good thing

Well, maybe not for the drivers who see 90% of their income earning opportunity automated away and all they get is the local driving on each end. Well, probably on one end -- the warehouse where the truck is loaded will be located where an automated truck can drive to it easily.


> who have been driving our cargo around for whole generations

What does that add to the argument? Is there some inherited right to work in a specific line of work? Also, how many generations can that be? 3-4?


Yeah, especially that self driving cars have been in the news for years now. Everyone know they will take over. There is plenty of time for people to retrain to something else. Unless they are lazy and don't want to...


Or you know, humanity is terrible at retraining people - that the costs of retraining are so high that the simple straight path idea everyone has is dumped on contact with reality?

Remember that the greatest advance to education was the internet.

You could now have good quality, always available, free, subject material, from reputed institutions, with interactive media, perfected explanations, testing and so on.

And despite all this - they have a 10% completion rate.

I have said this multiple times, and unless people are actively looking at this system they miss that re-educating and re-training people is extremely lossy and unlikely.

And now you have people out of a job, and they have to re-train, while also taking care of bills and family?

People tend to forget - our best training system is to take young humans with no responsibilities, mostly spare time and keep them in training centers for the vast majority of their day in childhood.

And even there we fail.

And finally - please come to India or any third world country. It should disabuse people of the idea of the fully autonomous self driving car.

The only way you get level 5 autonomy is by controlling the road and entry to it.

It's exceedingly easy to see a bunch of cows or protesters actively hamper fully alert human drivers.


Having grown up in an area where becoming a trucker is a common enough career path, I can tell you that the hours of work thing (which is really miles per work, but they correlate) lands some truckers in some sketchy situations

Many of my friends have told me they take those gas station "energy supplements" so they can stay up for 2-3 straight days to log more hours and make the trip.

Thats probably not healthy for their bodies, and certainly not safe for other people on the road, but if miles = money then its whats incentivized

If trucks could drive themselves with a passenger trucker to intervene in tricky situations, the passenger could have a much more healthy intake and sleep schedule


"so they can stay up for 2-3 straight days to log more hours and make the trip" - by the way, is that legal and actually happening in USA? Sounds like a horrible risk to everyone else on the road; in EU you're not getting on the road without a tachograph recording your driving & rest times, and driving more than 9-10 hours per day (including at least 9 hours uninterrupted "rest" eash day) would result in massive fines; if the cargo is urgent and you need to drive on then it's literally cheaper (as in, it actually happens) to fly in a replacement driver with a low-cost airline than to risk the fines.


No, it's absolutely not legal. The big difference is AFAIK, we don't require electronic recording. A lot of trucks owned by the bigger trucking companies do have electronic recording though.

The max per day is 11 hours, with a 70 hour cap for the whole week.

https://www.transportation.gov/briefing-room/new-hours-servi...


>trucks could drive themselves with a passenger trucker to intervene in tricky situations

this scenario has been mentioned a couple times in this thread, but if the automation isn't good enough to handle 100% of the journey 100% of the time, then it's really not good enough to allow the trucker to simply be a passenger. If the operator might have to take control in tricky situations, they can't be sleeping the rest of the time.


Well, but what if it can handle 50% of the journey 100% of the time, and that 50% is known?

IE, for the 100 mile stretch of road in the desert, when it is sunny, it works great, and then when it gets to the city it tells the driver to take over.


> If the operator might have to take control in tricky situations, they can't be sleeping the rest of the time.

A tricky situation for a computer is often not a tricky situation for a human. 99% of it will be the AI getting confused by things that are simple for humans (e.g interpreting detour signs).


that doesn't change my point though: regardless of how easy it is for a human to respond to, they can't respond if they're sleeping.


If trucks could drive themselves with just rarely needing someone to intervene in tricky situations (and not immediately - e.g. the "passenger driver might have to wake up) then it's going to be much, much cheaper to have the "tricky situation intervention" done remotely by a having a single on-shift driver ready to handle tricky situations for a hundred trucks.


hmm, Im not an expert but isn't your logic based on constant supply/demand? Once self driving trucks become mainstream could it be that truckers will do fewer long trips but more shorter trips which doesn't necessarily lower their income?


> Truckers get paid per hour of work, less work == less money.

Such a trivial and stupid reason to guide the future of technology and the human conditions. Of course people want to be home with their families instead of driving for weeks on end.

I find it dizzying that we think we can automate all of the matter around us to do what pleases us, but the matter of changing the socioeconomic configuration to match it seems a bridge too far. Come on now.

The economy (and society) will change, and this dude can stay with his family. Sounds great.


The people who are doing the automation, or own what is automated, don't want to change the socioeconomic configuration, it's working out very well for them.

> The economy (and society) will change, and this dude can stay with his family. Sounds great.

Given the productivity increases we've seen in the past few decades, this really hasn't been the case.

It is hard to stay home with family, when "home" and "family" means having many low-paying jobs to support both.


It is probably harder to automate the last few miles. The Uber depots in the video can be standardized, but terminal pickup and dropoff is legacy. That cramped loading bay in the downtown Giant Eagle in Pittsburgh is never gonna get much better.


Yeah, I imagine that bit about needing needing a human is a direct insult to their intelligence. It's just there to distract them until we no longer need it and they are entirely out of jobs.


> The first part can be trivially automated

Having done some cargo loading and unloading myself, good luck with that. It's a very heterogeneous environment and task.


I think it will be some time before the big shipping players all decide to switch to self-driving trucks. We’re still a decade away from complete vehicular automation.

> The first part can be trivially automated and the second is just stupid. Truckers get paid per hour of work, less work == less money.

You could always pay them a salary and benefits, and hopefully UBI would exist at that point. Of course it’s unlikely for Uber to do that, because their strategy has been the exact opposite.


weaving machinery is going to drive out of work all the textile workers!


No, no, that was off-shore child labor that did that.


I assume these self-driving trucks still have a human driver in them. So, which exact parts are self-driving? Is it like Tesla's autopilot where it does lane following, cruise control and keeping the distance on highways?


It's just adaptive cruise and lane keeping. Just like trucks already have now. The headlines to attract clicks on these are so infuriating...


It's not self driving. But it may be One of the first places that get paid real money(lets say UBER'S commission) for testing self-driving.


> So, which exact parts are self-driving?

Just enough of it that they can skirt the DOT hour rules about hours worked while still blaming the driver for any crashes or tickets would be my guess.

That's the obvious lowest hanging fruit as far as trucking is concerned. If you can automate enough of the long haul grind that the driver could possibly be sleeping then you have a massive competitive advantage over everyone else.

If you don't automate enough to be able to skirt the DOT time rules there's not much of a competitive advantage.


I don't think you can skirt them. Time on-duty counts against their daily limit even if they're not driving (the example given in an official guide is a trucker taking a bus to get to the truck), so they must still get a proper rest afterwards. So Uber can't count the "not driving but watching the road" as time off.


They're not going to be on duty. They're going to be sleeping in the sleeper, on paper at least.


And how will they blame a crash on the driver, when the paper say s/he was sleeping?


> Just enough of it that they can skirt the DOT hour rules

Source?


Source: plausible extrapolation based on Uber's storied history of violating labor laws to one extent or another in just about every market they've entered.


Come on, I'm not a fan of Uber and their shady history, but you can't just invent claims out of thin air based on your feelings.


What feelings? Do we seriously need to review Uber's conduct? We over here pretending their corporate culture has shifted to something approximating ethical behavior now?


No, I'm not trying to excuse Uber or say that they're some amazing benevolent company.

All I'm saying that I think, if you're going to make a claim like above it at least comes with a source or something relevant to the claim itself to point at as proof.

But look, if that's too much to ask for on Hacker News then fine.


I think you were right to call him/her out. Comments like that have no place here.

Level headed conversation with references and sources if possible are what I come here for and I think what the community expects.

The other stuff can stay on Reddit.


I think these companies (autonomous car/truck) understand that they need to transition these processes slowly over time. I feel like it's better to start with something only moderately different and iterate than try to completely reinvent the entire structure of such a large industry.

First they do this, then they automate more and more portions, at each step reducing the amount of manual, human supervision. It also allows the industry to more consciously reduce numbers through attrition rather than mass layoffs. Better to stop young guys from becoming truck drivers and have the automation increase over 10-15 years until it just really isn't a career. The only people that will be affected are in the 25-35 age range now who have already taken the plunge, but at least they get some time to consider other options.


This is a woefully optimistic viewpoint: when the technology is ready there will be layoffs. IMO.

These companies aren’t slow-playing their hand to avoid layoffs, and it’s a race to see who can perfect the technology first.


No doubt, but I think it's just the nature of complex systems. You could invent a drug that cures all diseases and maladies, and there would still be practicing doctors for 5-10 years while the cost/distribution/red-tape was worked out.


> I think these companies (autonomous car/truck) understand that they need to transition these processes slowly over time.

They understand pretty well that they need to transition faster than their competition. Given that, they'll do it as slow as their competitors allow, because it's investment and they don't like doing that quickly.

How fast that means is anybody's guess. But they can't just stop and wait.


I feel like they re-invented the train. :-)


Trains are much less flexible. They have more constraints on their schedules, they have to share tracks, they only go certain places, you have to stop the entire train with all its freight in order to add any more freight, the unloading and loading points are much harder to scale.


Laying rail is fraught with property rights to build the track.

Roads are like rails that are already paved everywhere (with the significant addition to climate-impacting gas emissions).

Rails are easier in Europe because less of the land is owned by private interests making eminent domain easier.


Without the rails.


Will Uber pay a bigger portion of the road repair taxes? Especially considering truckers already pay a disproportionately lower tax compared to the impact they have on the roads.


In the United States trucks are already taxed at higher rates because of their weights. That's why we have weight stations scattered throughout our highways though automation and technology have reduced their need.

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


I think there's still an argument that it's not enough, as road wear varies with the 4th power of tire load. Ultimately time is another factor in road wear and breakdown, so utilization rates matter, but on balance freight traffic is subsidized.


I'm sure Uber will come up with some bullshit reason why they don't have 'trucks' and don't have to pay taxes.


This represents a fundamental misunderstanding of the rationale for taxation.

There is a societal benefit to efficient means of transportation. You, along with most other people, are reaping the benefits of truckers using roads to deliver cargo.


Not including externalities--such as road maintenance--distorts the market. Better to capture them all so that different forms of transportation can compete "fairly". That is, without hidden subsidies.


The rationale for taxation has changed over time. We are no longer on the gold standard, so the reasons for taxation are different. But overall the reason for taxation hasn't changed, which is to maintain the demand for the central money supply. This of course is the rationale for sovereign entities with their own money supply like the United States. The rationale for taxes for towns, counties, etc is completely different.


Do they? I mean, trucks are hardly gas efficient, and aren't most roads maintained (supposedly) via gas taxes?

The question, I suppose, is how much harder they are on the roads relative to their fuel efficiency.


Fuel efficiency decreases generally relative to the surface area of the front of the vehicle, where road damage increases with the mass of the vehicle. So a long, heavy vehicle like a truck is indeed harder on the road relative to their fuel efficiency.


  > road damage increases with the mass
Road damage actually increases proportional to mass^4 - trucks cause exponentially more damage than cars.

https://streets.mn/2016/07/07/chart-of-the-day-vehicle-weigh...


> I mean, trucks are hardly gas efficient, and aren't most roads maintained (supposedly) via gas taxes?

The gas tax hasn't been increased since 1993 and accounts for less than half of road funding.


Perhaps on a related note, our roads are in embarrassing condition.


That sounds like a problem with the trucking industry, not Uber. So the trucking industry as a whole should be the thing being taxed more.


Isn't HN particularly fond of mentioning that one of the pillars of Uber's business model is to externalize costs? So no...


When trains become competitive with road transport, maybe we should move more to using trains. Until then, we won't, and we shouldn't. (If it's a matter of externalities/hidden subsidies, then the costs of these externalities must be brought to the entity causing them, e.g. via taxes.)

I don't know why (in Germany) a long-distance bus is much cheaper than a train (from/to the same city), but it is. We shouldn't be trying to shift traffic to rail unless there's a good reason beyond "it feels good" for it. If it is better for the environment, I'd like to see a "citation" (not in the form of research, but in the form of well-justified taxes/tolls/... that transform this into a matter of cost).


If trucks had to pay the full cost of road wear, then trains would be a lot more attractive for long hauls -- road wear is proportional to the 4th power of axle weight [1]

An 80,000 pound truck has 8X the axle weigh of a 2000 pound car, so should be paying 4096 times more road taxes than the car.

[1] https://en.wikipedia.org/wiki/Gross_axle_weight_rating#Impor...


Sort of. It looks like a car does about 4x more damage per axle because of using a single narrow tire[1]. If you compare a pretty typical 4000 pound car to that maxed-out truck, you would then only get a 64x difference (maybe closer to 100 depending on weight distribution). Oh and then multiply by the number of axles.

It would be interesting to look at the costs and savings of 6 or 8 axles in the back. There's plenty of room to fit, and you could drop wear by a factor of 10.

Let's also not forget that the truck probably has over a hundred times as much payload as an average car.

[1] http://www.nvfnorden.org/lisalib/getfile.aspx?itemid=601


In the US, trains are for long distance and, in fact, handle a significant portion of the long distance transport that would otherwise be handled by trucks. Most trucking is (relatively) local as it is.


Railroads have monopoly pricing power, this opens the door to competition


It opens the door to some company buying up the roads, and then you're right back to where you started.


Private highways are already a thing


Snow Crash here we come.


I am not sure that I want a company that is most famous for breaking rules, skirting regulation, and stealing medical records of it's victims operating large freight trucks on public roads.


This is a good point, but it can no way be worse than what we have now.


Notice that you can extend this model to passengers and long-haul bus, train, or simply self-driven vehicle transport, in all of which cases Uber is well positioned.


Should be a fun day when people realize they can pretty easily hijack automated trucks.

I assume you could either get in the back of the thing while it's moving, or simply slow down in front of it and cause it to stop. Unload your newly acquired goods and be on your way!


What makes you think they would be easier to hijack?

Part of the ease of hijacking truck is that you can drive off with it. What if the trailer locks its wheels and bricks itself, if it determines it's being hijacked, all while recording 360 video and sending immediate alerts to management and law enforcement?

Certainly better than a driver afraid for his life.


Think burglary vs home invasion. It's not just about easier, it's about it being a different threshold for whether it's worth it. The presence of other people drastically complicates crimes, and makes it much more likely that you'll end up dead or in prison for decades.

Also, all the countermeasures you mention aren't free or without issues, and could be in place now, but aren't. Why not?


In general, the real bottleneck with theft at scale isn't the actual theft. It's turning the stolen goods into money afterwards and disposing of the evidence. Items being slightly easier to steal doesn't do much to make them easier to fence.


You can do that now. Most truck drivers aren't going to defend their cargo with a gun.

So what's stopping people from doing that now?

Note: an automated truck can probably dial 9-1-1 as easily as a human.


Can't you use the same strategy for human driven trucks? Except that the human driver is a lot less likely to be uploading a live video stream back to the corporate office.


Assuming the truck driver isn't heavily armed or willing to endanger themselves by wrecking an 18 wheeler, there's nothing really stopping anyone from doing that today.

You seem to be working under the assumption the nobody has ever hijacked truck up to this point, and that for whatever reason self driving trucks would be left unlocked with no security mechanisms in place whatsoever.


Can't we get this technology nailed so well that hopping in a self-driving car for your day-to-day errands is something you do without a second thought, BEFORE we start sending out 30 ton missiles as beta tests?


I agree with the sentiment, but guiding a large missile down a straight highway with human oversight has many fewer variables than navigating smaller missiles through crowded urban areas.


Plus given the amount of trucking on certain corridors it might be possible to get some lanes ‘roped off’ for automated trucking only, further reducing the risk.


Obviously not, since carrying humans on arbitrary dynamic routes for day-to-day errands is a much harder task than driving a known route between two fixed, well mapped points, and navigating remote roads/streets is more difficult than driving on highways.


The missiles involve more economic productivity.


Does the self driving truck have any kind of security? Given that there is a "do no harm" program embedded into the driving "subroutines", is there anything that prevents thieves from essentially surrounding the truck and robbing it in the middle of nowhere?

Edit: To clarify I mean - in a hypothetical situation where it is ever driverless, couldnt 10 of us surround it, while I walk in and commandeer it, disable GPS and drive off ?


I mean, if 10 people surrounded a truck with a single human driver, they'd probably be able to drive off with it too.


A future 'true' driverless truck would likely not have any human controls. Or the controls would be computer locked and would require a code to be activated.


I imagine these trucks would phone home immediately and warn an operator that they had unexpectedly stopped. And I'm sure they continually log their GPS location with a server somewhere. They likely have some level of video feed being logged as well.

You probably won't be able to steal the truck and get away with it any easier than you could a human-driven truck.


The movie Solar Crisis already covered this:

https://www.youtube.com/watch?v=i729rZWTmDo#t=30m22s


You can already do that to a normal truck. It's not really a problem. It'll be rare enough to fall under the umbrella of "that's what insurance is for"


What prevents 10 of us from surrounding a real truck, walking in, forcing the driver out and driving off?

https://xkcd.com/1958/


I guess we are presuming that drivers wouldn't simply step on the gas in that kind of situation.

Also I suspect the absence of a human being in the situation would psychologically make the theft feel "less wrong" to many would-be criminals.


> Uber has a video depicting this journey, which took place staring in then midwest and then via short haul to Sanders, Arizona, where it was loaded onto an autonomous truck and then transferred to Topock, Arizona, where it made the switch to another human-driven vehicle. This is a big step towards commercialization of Uber’s autonomous truck tech, and it seems like it puts it ahead of some competitors who want to do similar things, including Embark trucking. It’s also the first we’ve heard about Uber’s self-driving truck business in a little while, so it’s good to see Uber’s continuing to make progress and devote attention to this issue.

Uber's main line of business i.e. on demand car transportation is loosing money. They tell a story that they fix it with self driving cars which incidentally would require gigantic additional funding. Now they have expanded into self driving trucks that do already some services with some capacity in the name of their own surely significant trucking business for some customers that call for them through their app. This is the real deal, I know because there is a video.

Musk heaps one money loosing business over the other but at least they all make sense and are different. With Uber it was doubling down and now tripling down. Not impressed.


I would be interested to know how would these autonomous cars/trucks behave on non-linear roads, with potholes, adverse conditions, etc

What kind of problem are these companies trying to address that hasn't been addressed yet?

Will these trucks be a replacement for trains? I can see the later ones being way more easy to automate (e.g. London Overground)


That’s probably why Arizona is the test bed right now. Predictable weather patterns, long stretches of flat, open roads, low traffic, and legislative openness. optimizing 80/20


I'd expect they'll make potholes much worse, with all trucks driving exactly over the same spot repeatedly, until they learn to avoid the pothole. Perhaps Waze-style reporting will fill the gap.


Don't you think they'd be able to take that into account with the software? Large trucks already put a huge amount of wear like that on highways (I've seen plenty of ruts from trucks in asphalt roads). Technology like this is an opportunity to actually improve the wear patterns.


Without wanting to sound too cynical: Why would a private company care about the wear patterns of public highways?

Unless it's affecting the operation of their fleet, in a negative way, they don't have any reason at all to improve anything about it.

Especially considering how this does not look like a trivial problem to solve. Wouldn't something like this require an additional sensor suite to check for more detailed road conditions? What about the future, when other companies start doing the same, wouldn't that require some level of cross-company cooperation to make sure trucks take somewhat randomized lines on the street, to reduce wear?


You're referring to the tragedy of the commons I believe. So the simple answer is that they might not, and so perhaps there's a role for government oversight/requirements similar to safety requirements. It's not a stretch to consider this a safety issue, if only for the other human-piloted vehicles.

But I think it could be in a private company's interest too. Better road conditions would be better for wear and tear on their vehicles. Lower maintenance costs on the road should result in lower tax burden for the company (assuming company pairs fair share of their taxes which admittedly may not be true).

Cost of software improvements like this scale well (ie cost is distributed over time and customers well). In fact the vendor of the vehicle could present it as a competitive advantage over other vendors that don't address it in any way.

I wasn't thinking the solution would be based on sensing the build-up of lane wear/ridges, I presumed the solution would be something along the lines of varying their lane position to avoid the specific case of always travelling in exactly the same spot in the lane. With some randomness I don't think it would take inter-company cooperation. Just the idea that if the same vehicles were travelling the same route, that the wear on the road would be reduced by some amount. Obviously this is a lot of armchair architecture-ing from some random front-end UI developer, but I think it's not unreasonable to think along these lines once they reach the point of optimisations. Fun to think about anyway.



I don't mean to be a downer...but is there really anything novel here? I understand that this is a first step, but I feel like Uber could have significantly reduced the manual labor required (e.g. to disconnect/connect a trailer). It seems that the load (mental/physical) on the "driver" of the self-driving semi will be nearly equal to a driver who has cruise control.


I am not sure how much of that equipment is legislated. The air fittings for example are of a type I have never really seen elsewhere. They have been standard in the trucking industry as long as I have known (Oldest truck I have been around was from the 70's)

While you could develop a way to robotically latch those it seems like it would be easier to redesign the fitting.


You’re right it’s not that novel. Volvo has self-driving* truck convoys going back to the 90s I think.

I think for long-haul trucking it’s certainly pretty hard to stay fully tuned while driving for such long times. And for truck companies, they can save tons of money on labor when they don’t have to pay the driver for the hours on the road and just for the manual disconnect/connect docking. Maybe they can make it fully automated, but why haven’t any trucks done that already? It’s an easier problem to solve than self-driving.

* by convoy, I mean that one truck up front was driven by a human and then a few were automated to draft from directly behind to just follow the truck in front of them.

EDIT: Here's a source for the 90s, where this paper had a few authors from Volvo.

from: http://orfe.princeton.edu/~alaink/SmartDrivingCars/ITFVHA15/...

>Additionally, different applications of automated driving for trucks have been demonstrated across the globe since the 1990’s. The focus has been on proof of concept for truck platooning due to the foreseen fuel economy and traffic flow benefits.


Truck companies avoid this kind of thing because it's expensive to maintain pools of short-haul drivers at either end of the transport link. If you are willing to assume those costs/risks, you can already send freight by rail, which is an order of magnitude more efficient on fuel, and sometimes even a little faster.

The reason you pay for long-haul trucking is that you want to have the driver _already on hand_ for the two short-haul trips at either end of the journey, navigating complex issues and handling cargo loading. The downside is that you have to pay the driver for the over-the-road portion. The upside is that there's zero chance of a driver not being available, and zero chance of the driver being left idle waiting for the long-haul transport to arrive.

The only novelty here is that Uber is the one paying the pools of short-haul truckers on either end, and assuming the risk that they will sit idle. (They are, of course, not going to tell us if it is unprofitable!)


The US already has one of the largest freight rail networks in the world (at the expense of having a totally shitty passenger rail network; although that's starting to change with Florida and California's efforts).

I wish more effort would be put into expanding freight even more so you could get the same speed and efficiency for shorter hauls, mail and package delivery. It seems like that would be way more worthwhile than automated trucks. It's substantially more trivial to automate freight trains as well. Passenger trains in Singapore, Malaysia and parts of London are already fully autonomous with no driver at all.


The U.S. used to have extensive short-haul rail, even inside cities. That's how cargo moved around back in the era before inexpensive motorized trucks.

I suspect the reason it disappeared is that rail sidings are large and expensive compared to truck loading docks. You're not going to have ten sidings on a property as easily as you stick ten bays in a loading dock.


> Truck companies avoid this kind of thing because it's expensive to maintain pools of short-haul drivers at either end of the transport link. If you are willing to assume those costs/risks, you can already send freight by rail, which is an order of magnitude more efficient on fuel, and sometimes even a little faster.

Speaking only from the US perspective, I agree with most of this. But there are a couple things you're not taking into account:

1) Delivery variability for rail. While rail is much cheaper in terms of cost per ton mile, good luck getting an accurate estimate of when it will be delivered. Midwest to Pacific Northwest I've seen variability of a week or two for going halfway across the country. If they're delivering to you and are multiple days late, they still expect you to drop whatever else you had planned and get their rail car unloaded on a relatively short turnaround or they start charging fees.

2) Rail requires quite a large scale. Many factories aren't shipping out more than a few semis per day. Getting to enough to fill a couple rail cars would significantly increase the inventory levels.


Since the 90s?? That's very impressive if true. I remember reading about a proposition to do this for cars years ago, before any talk of self driving vehicles. A car would lock in behind a professional driver vehicle such as a truck or bus and match the turning and braking. I didn't think the tech existed though


The 90s wasn't exactly the stone age, there was plenty of self-driving vehicle work taking place. For example:

https://www.smithsonianmag.com/history/the-national-automate...


I edited the original comment to include a source. Serious talk of self-driving cars was around by then, as they had already successfully used auto-pilot in planes, it wasn't that infeasible.


It seems like truck convoys are a less efficient reinvention of trains. I'm still surprised they didn't take off in the States, though, since our rail infrastructure is pathetic. Maybe it only works now because Uber can burn $1B a quarter.


Why do you think our rail infrastructure is pathetic? It's the largest freight rail networks in the world.

It sucks for passengers but that's not relevant to this discussion.


You're right -- it's better than I thought, though it's still in the middle of the pack when measured by track per person or track per area:

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

I guess I'm biased by seeing all the abandoned rail, since it's well down from its peak.


> it's still in the middle of the pack when measured by track per person or track per area

Both of those seem like somewhat irrelevant metrics. Why should a more densely populated country be penalized for having larger clusters of people compared to (based on this data) Canada or Russia? On the other end, why should a country with large swaths of unpopulated land cover it with rail to improve its track per area metric, which densely populated countries seem to do well on?

It may be more useful to consider the efficiency and cost of freight rail, as well as the share of freight transit (per unit of distance, for instance) that is on rail.


They're not perfect, but they seem useful to me: roughly how much stuff per mile is on your rails, and how far do you have to travel to reach it? Do you have some cost per mile figures to share?


Yeah, the US rail system is the best in the world for freight. In other countries, passenger rail has priority, so the freight service suffers a bit. In the US, it's the other way around.


That's also why Amtrak is such garbage compared to any civilized country.


Check out the usage statistics.

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

Most tonnage shipped by rail in the world, with 10x more than the EU.


My guess is the point of this video is to paint a picture of AI as really pretty boring and nothing that much different for the driver.

This is a tactic to avoid pushback from the public / truck drivers.

This is an attempt to control the narrative on what autonomous shipping is going to look like.


What really bothered me at the end of the commerical was the fact that the person could be in home in time for dinner.

Unless we have some form of UBI, where is that person going to source the funds to pay for his dinner?


What I find truly interesting of this thinking is that technological advances in the future are cataclysmic, but in the past are necessary.

It never becomes something like "lets destroy all the farming machinery, so we create more jobs plowing the land".


Oh, you can rest assured that we have enough cataclysms in history to make the case.

There's that entire "let's destroy all machinery that is taking your jobs" movement that gave a name to the "avoiding new things" attitude. All that in a period when people were choosing between literally dying of huger unemployed or of extenuation at work.

Just because the cataclysm is temporary does not mean it didn't exist.


Would you advocate to destroy all the machinery those people clamored against today?


Of course not. But some welfare would have saved a few million lives, greatly enriched any country (although people didn't know that by the time), and removed one or two world wars from our history.


You got some leaps of faith there.

Capitalism was basically born by the application of technology disruptively like it did. Saddling the creative forces with the consequences can easily prevent the change from happening at all, and in the end, it is better for the whole.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: