Sep 12 2017

The Safety and Ethics of Self-Driving Cars

Google-SDCGermany just came out with their first regulations for self-driving cars that address how they will be programmed with respect to safety. Specifically – what should the programming do if harm cannot be completely avoided and it has to decide between the lesser of two bad outcomes? Germany is the first country to come out with such regulations, and therefore sets the example for other countries who will likely follow.

Here are the key elements of their decision:

  • Automated and connected driving is an ethical imperative if the systems cause fewer accidents than human drivers (positive balance of risk).
  • Damage to property must take precedence over personal injury. In hazardous situations, the protection of human life must always have top priority.
  • In the event of unavoidable accident situations, any distinction between individuals based on personal features (age, gender, physical or mental constitution) is impermissible.
  • In every driving situation, it must be clearly regulated and apparent who is responsible for the driving task: the human or the computer.
  • It must be documented and stored who is driving (to resolve possible issues of liability, among other things).
  • Drivers must always be able to decide themselves whether their vehicle data are to be forwarded and used (data sovereignty).

This all makes sense to me and I don’t see anything overly controversial. Prioritizing people over property is a no-brainer. Treating all people as of equal value also seems like the right move. This is because you could not individualize such decision – only treat people demographically or as part of a group. This would be too ethically fraught to be practical.

The rules assume for now that “driverless” cars have the ability to drive themselves, but still require a licensed capable driver to take the controls when necessary. In fact, the report discusses the possibility that if the programming encounters a decision dilemma with regard to minimizing death and injury, it may turn control over to the human driver at that time to make the tough decisions.

That was the trickiest part of the recommendations, in my opinion. I certainly wouldn’t want the car to suddenly turn over control in an emergency situation. The driver would be hard pressed to react quickly enough to produce a superior outcome to the programming just doing what it can. The report does say if the driver does nothing it would brake and come to a stop as a default. I suspect we’ll find that the default response of the car is likely to be superior to the outcome of giving the human driver sudden control.

The report also explicitly recognizes that automated cars are safer than human-controlled cars. This is already true, and they will only get safer. After 1.8 million miles of driving, Google’s automated cars were in 13 fender-benders, 100% caused by other drivers.

Further, having automated cars on the road will increase safety for everyone – not just those in the automated cars. The higher the percentage, the safer our roads will be. Germany concluded this creates an ethical imperative for governments to facilitate self-driving cars. Perhaps one day they will be mandatory.

The reasons for this enhanced safety are obvious. Driving safely requires constant attention and the ability to react to a sudden hazard with little warning. Humans are terrible at constant attention. We can become tired, distracted, and lose focus. We can become confused by complex intersections or road signs, or blinded by the sun. We may decide to drive while impaired, either because we are sleepy or inebriated.

Computers do not get distracted, do not lose focus, and are never fatigued. They are simply better drivers than humans. This fact alone is creating great pressure to adopt automated vehicles as quickly as possible, now that the technology is here. It seems like this tech will be like smart phones – in a very short time they will become the norm. With cars, however, a “short time” could be 20 years. This is because people tend to hold onto their cars for a long time – the average age of a car on the road in the US is 11.5 years. So even if all new cars purchased were self-driving, it would take a decade to replace most of the cars on the road.

But there are some areas where self-driving vehicles may take over more quickly. Trucks are one example. Shipping requires driving long distances, a much better job for a computer than a human. There are lots of trucks on the road and if they had the enhanced safety profile of a self-driving car that would make the roads much safer.

Taxis are another example. Uber has already demonstrated that the old taxi model is outdated. The new model of using an app and and algorithm to match driver and passenger is much better. Now make those Uber cars self-driving and the roads may quickly fill with driving as a service. This may significantly decrease the need and motivation for even owning a car for many people.

So even if there are a lot of old traditional cars out there for the next 20-30 years, the roads may disproportionately be occupied with automated vehicles fairly quickly.

This is a good thing not only for safety, but for efficiency. Automated vehicles are also likely to be more energy efficient. They can be programmed to optimize acceleration and braking to minimize energy use. Again this is something that humans are not very good at. Cars that give real-time feedback to the driver about energy use do improve fuel efficiency, but computer algorithms could be optimal.

Further, an integrated system could theoretically reduce travel times. Traffic could be streamlined and optimal routes calculated. Also, with Uber type services, you won’t have to make round trips to drop people off.

Once we get to the point that vehicles are fully automated we won’t need a human driver at all. This could add further efficiency. This would reduce the number of people who need to be in the vehicle, assuming the driver is only providing the driving service (like driving a truck or dropping someone off). That is 150 pounds or so less in the car, which improves fuel efficiency. This would have the added benefit of saving time for all those unneeded drivers.

There really is no practical downside to changing over to self-driving cars. This is a huge win for society. One potential subjective downside is the perceived loss of freedom and the fun of driving. That factor may linger for a generation or two, but eventually driving a car will likely become like riding a horse – a pastime or sport rather than a necessity.

The German report mentions the vulnerability of hacking. This is clearly a risk. They recommend that measures be taken to make automated vehicles and their support systems as secure from hacking as possible. This is likely going to be the biggest issue of safety, and we should definitely be spending a lot of resources figuring out how to bake in security from hacking right from the beginning. All such cars should, at the very least, have a fully manual override. The human driver should always be able to take full control in a way that is hack-proof.

But that is not enough. Self-driving cars will need extreme security from hacking. To me that is the only big question mark – how secure can we make them?

Germany’s new regulations are a step in the right direction, and hopefully will motivate other countries to follow suit. This is going to be a relatively rapid technological change, and it’s best that governments get out in front of it thoughtfully.

54 responses so far

54 thoughts on “The Safety and Ethics of Self-Driving Cars”

  1. There is one downside. Entire professions and industries will be eliminated or at least severely downsized. Driving professions will be largely eliminated. Auto body repair employs a lot of people, and buys a lot of parts. Cars will last longer, so demand for them will drop. A lot of health care is for people injured in auto accidents. I’m sure with a bit of thought numerous other examples could be thought of. While I agree overall it’s a great thing and the wave of the future, I hope governments give some thought to the economic changes the switch to driverless cars are going to cause.

  2. united – I don’t see that as a downside. That is the creative destruction of the market. Anything which adds efficiency to how we do things reduces jobs – that is kind of the point of greater efficiency, we do more with less. That is a good thing overall.

    But yes, that means we need to make sure our economy is also optimized for such changes, meaning that workers have the opportunity to migrate to other jobs and other industries, that there is an infrastructure for retraining and support for people during the transition. Absolutely. We should focus our efforts there, and not perceive technological advance as a downside.

  3. Kabbor says:

    The problem of automation taking over jobs is always a challenge because the beneficiaries of the new technology are the users of the product or service, and the remaining providers of the product or service. The ousted labourers do not benefit from the efficiencies created by their loss of employment and this is tricky because well paid labour typically has a high education and/or training requirement.

    A quick question about hacking: Is the fear that these vehicles will be hacked remotely, or that people with ill intent will physically connect to the computer in the vehicle and hack it that way? It seems to me that the computer would not need to be wireless accessible, and making it so that you need the key in the ignition before it accepts computer input might go a long way toward making the cars challenging to hack. You’d need to be in the car and have the car key in this scenario. I’m not a computer programmer, so I am way outside my field of expertise here. Is it that difficult a problem?

  4. Daniel Hawkins says:


    The problem is that cars access the internet, for navigation/traffic updates, Spotify, software updates from the manufacturer, whatever, which creates an opportunity for infection by a virus/malware. So even if the car still required a key, it could change the driving behavior—for instance driving away to a remote location at high speed and holding the occupant ransom. And even the requirement for a key could be bypassed if there was a severe enough security hole, so the car could simply be stolen.

    There’s no such thing as a hack-proof system.

  5. trampuspi says:

    Maybe safer… Sure, it works in well controlled experiments, until you factor in millions of people not keeping up with maintenance, not keeping them squeaky clean and in perfect working order. Not to mention people tweaking the software with apps and you-tube advice. Then there is cost, only upper middle class and up will be able to afford to have a car, as a good portion of Americans can not afford brand new human driven $40,000 cars.

  6. daedalus2u says:

    Self driving vehicles don’t “need” to be connected to the internet for updates. They could periodically go to a secure location for updates via direct connection, such as when charging and/or for maintenance. They will need to charge themselves every few hundred miles.

  7. Kabbor says:

    Ah fair enough. I don’t have a car with any self-update features or navigation software so I did not consider them as being points of entry. I imagine that if the manufacturers were determined about it on the software side they could make the updates require a physical USB key or something, but it would not solve the potential navigation input issue. There you will need Hollywood-style white hat hackers pounding away on a keyboard in front of an incomprehensible screen in order to defeat the efforts of the bad hackers.

  8. chikoppi says:

    I admit to a significant apprehension with respect to self-driving vehicles. In particular, mixing driven and driverless vehicles in crowded conditions. I do think driverless technology could be a boon to vehicular safety, were it to be applied as driver assistance systems.

    However, I could envision a freight or mass transit system that operates separately from “civilian” traffic. This would entail fully separated roadways or lanes. Such automated systems could operate between hubs and terminals (that also serve as charging stations). At the hub cargo might be transitioned to smaller driven vehicles (delivery vans and taxis) that serve the last local leg of the journey. This would be made more efficient if the containers were modular, to avoid having physically transfer the contents between trailers.

    There are many potential applications that might require thinking about infrastructure differently.

  9. Nareed says:

    I wonder how self-driving cars will impact short-range air travel.

    Suppose you need to be at mid-week at 9 am somewhere that’s a 10-hour drive away. If your car can drive you there in reasonable safety, you may choose to drive rather than fly, especially if you can sleep through much of the trip

  10. You mentioned computers not getting “fatigued,” which brought to mind something I hadn’t considered before. Cars obviously do deteriorate over time and can have issues that make driving unsafe. If a self-driving car can self-diagnose (some expansion on the engine codes in use now), will it be able to prevent its owner from taking it out on the road? Should owners be able to override such decisions? Either way, how does liability work?

  11. Dan Dionne says:

    Steven, you make great points about the benefits of self-driving vehicles, and I’m totally on-board. I don’t think the economic consequences for truckers should be underestimated, however. There are 3.5 million truck drivers in the US, and growing. Displacing this much labor is a genuine downside if these workers don’t have any other options. I think we’re looking at a labor shift that will be faster and more complete than what’s happening to the coal industry.

    Once self-driving trucks are cheap enough and available in large numbers (within fifteen years?), carriers will require over-the-road truckers to have self-driving capable trucks. Truckers who own their trucks will be screwed, unable to sell their now-worthless trucks. Big freight carriers such as FedEx and R&L will transition to new long-range fleets. Drivers will still be necessary to mind the machines, but wages will plummet as they spend less time actually touching the controls. As their safety record grows, carriers will push to legalize driverless, fully-automated trucks for regular routes between distribution centers. Smaller local routes will be next. Local commercial and residential deliveries will continue to need humans, but wages will flatline.

    So we’re probably looking at millions of truck drivers who will lose their livelihoods. Unless as a society we recognize this ahead of time, and provide awareness and re-training options, these people are screwed. Even then, some will continue to hold on out of stubbornness (I’ve known more than a few truck drivers, haha). I’m all for transformative technology, and excited for the self-driving revolution, but truck drivers have to be prepared before their skills become obsolete.

  12. SteveA says:

    Dan Dionne/unitedcats1957

    Easy solution.

    The ‘real’ truck drivers will just have to rebrand themselves as ‘organic’ and charge three times as much.

    Organic produce, delivered by an organic driver…Gwyneth will wet herself.

  13. fbrossea says:

    There are interesting arguments to be made. I think you can draw a parallel between automation and globalization because the end result is the same. The majority of economists agree that by many metrics, globalization is more beneficial to more people than harm. However, on a microscopic scale, it doesn’t ease the burn of losing your job in an outsourced field. The government has attempted to ease the transition through the “Trade Adjustment Assistance” program which pays for the retraining of workers who lost their jobs due to outsourcing. It is not perfect though, and is plagued with under staffing and imperfections such as forcing 59 years-old workers to attend trade school. TAA info here:

    Regarding constructive destruction, unfortunately some economist believe that now the construction isn’t keeping up with the destruction because there has not been the same surge in productivity with recent innovation as there has been with industrial innovations:

    It seems that the idea of a guaranteed universal income will need to be explored, but there’s an argument to be made about whether political pressure will make this happen before, or after many people lose their jobs.

  14. DanDanNoodles says:

    I am eagerly awaiting the arrival of large numbers of driverless vehicles. There’s a problem, however: once they reach a certain density, human drivers will almost certainly have to be outlawed, and that isn’t going to sit well with a lot of people.

    The reason is because human drivers will take advantage of the defensive nature of automated vehicles to become even more aggressive than they already are. We’ll see — even more than you do now — humans swerving dangerously through traffic, confident that the AVs will get out of their way. And there will absolutely be people who do it simply for the “fun” of watching the AVs scatter. That is going to happen. And the people who are most likely to be abusive in this way are probably the ones who are also most likely to not want to give up their “freedom” to drive.

    In the end, except for some professional classes of drivers — i.e., police and emergency vehicles — it is impractical to have humans and computers driving together. The temptation to abuse the system is just going to be too strong.

  15. Dan – I don’t think I am underestimating the impact on truck drivers. But, what you outline is probably close to what is going to happen. Only draconian and regressive laws will stop it, and we should not enact such laws. Let the innovation happen.

    We can deal with the consequences for displaced workers, if we have the political will to do so. First, we can see what is coming, so truck drivers should start to make plans now. Make sure they know. Don’t feed their denial with any talk of protectionism. No one is going to save their job. They should prepare for the future. They have 10-20 years to adjust probably. Start doing it now.

    Further, there are lots of ways they can be helped, in addition to what I mentioned, there can be a truck buy back program. Perhaps some of the trucks can be retrofitted to be autonomous (I don’t know if that would work, but it’s worth exploring) or just recycled.

    It’s the same as with the coal industry. It’s regressive and harmful to promise coal workers that you will save their jobs. It’s only delaying the inevitable. Far better to enact programs to help them transition to other industries.

  16. Daniel Hawkins says:


    Tesla already does over-the-air updates, but even if other manufacturers eschew them, the charging stations/secure locations are still points of vulnerability. Even if the updates have to be loaded via USB or other physical port, that still doesn’t eliminate the risk. Someone that is sufficiently motivated could gain access to the secure location via credential theft, impersonation, bribery, blackmail, etc.

  17. DanDanNoodles says:

    One thing these guidelines do not address is the question of whether an automated vehicle should protect its passengers’ safety over that of others. I would say “no” but I think a lot of people would be uneasy with the idea that their car might sacrifice them to save someone else.

  18. tb29607 says:

    Regarding current commercial drivers, I would add that in addition to retraining current drivers, we should also initiate serious reductions in the number of new drivers being trained.

  19. MosBen says:

    While overall I agree that the technology is coming and will be clearly superior to human drivers, I do think that Steve is underestimating the cost of lost driving related labor. This is kind of a different topic, but I do think that increased automation, like automated vehicles, have the potential to simply displace a sizable portion of the population without a comparable growth of new jobs, or at least new jobs that are actually available to them. A person that has been driving long haul trucks for 10-20 years isn’t realistically going to go back to college to study computer programming, or some other similarly high tech field. It’s easy to say that we’ll just have to summon the political will to help people through the transition, but our political system isn’t doing a great job of addressing societal needs at the moment, so I don’t see any particular reason to assume that it will get better soon.

    As for the technology, something that I’ve been thinking about for a while is how this technology could benefit mass transit. As people who have used Uber or Lyft may know, there is a ride-sharing feature where you pay a bit less for the fare and the software routes the driver to pick up another user that’s near-ish to your route, allowing the driver to take two people to two different locations more efficiently than if they had to be taken separately. It seems like this could be applied to mass transit rather effectively. Rather than have 50-person buses that travel along fixed routes that mainly travel along arterial roads we could have large automated vans that dynamically plot routes that pick people up where they are (or near where they are, maybe at pre-set pick up points) and take them to where they need to go (or near where they need to go). You’d never have to change buses , at least not within a certain geographic area, and because you’d be traveling with fewer people who were all going sort of close to where you were going your rides would probably be faster. You’d also never have the annoying experience of just barely missing your bus and having to wait an hour for the next one to come by.

  20. JimV says:

    I’m retired. Life without some kind of work is not enjoyable, even if you have enough money to live without working. The work does not have to be useful or very meaningful as long as it is interesting and provides some challenges, but without something to do (or to procrastinate from doing) there doesn’t seem to be much point to life.

    I think self-driving car technology is over-hyped and under-costed (it requires a lot of infrastructure and can’t handle all the tasks a driver may have to do), but it will work for some routine tasks, and will remove another source of work.

    In science-fiction, people don’t have to work in future societies and find hobbies instead. It is hard for me to see how those societies will evolve from the present, capitalistic ones. The direction I see us going in is for the 1% of rich people to let their money (usually inherited) work for them while they do hobbies, and for the rest of us to starve or compete with machines for the ever-smaller supply of work we can compete with them for.

    Climate change isn’t going to fix itself (in a way we would like) if we keep ignoring it and leaving it for future generations. Watching the supply of work diminish without any plan for what the unemployed workers are going to do may not be a good idea either. (There was a time when new technology created as many new jobs as it made obsolete, but I don’t think that is the case any more.)

    To modify an old saying: first the machines took away jobs from people who didn’t have jobs like mine, which machines couldn’t do, and I did nothing. Then the machines came for my job.

    Yes, it’s all part of nature and if Fermi’s Paradox doesn’t get us that way it will get us another way. Some ways seem like they could be avoided if we wanted to and tried hard enough, though.

  21. RC says:

    @Daniel Hawkins

    “There’s no such thing as a hack-proof system.”

    Currently, no. Theoretically, absolutely possible.

    The real problem is that even after the waves of online breaches, stuff with the credit bureaus, rampant id and credit card theft, etc, security of technology is very low on companies (and our gov’t)’s list of priorities. New features outsell everything else, and thus their development get the majority of the resources.

    We’re still using email for god’s sake, despite it being the security of equivalent of screaming something to somebody in a crowded library – when there are secure alternatives available. People just don’t care about security at the expense of anything else (with regards to technology) – which is strange, because most (Americans atleast) seem to be willing to give up an awful lot for perceived physical security.

  22. Ian Wardell says:

    Steven Novella said:
    ||”The report also explicitly recognizes that automated cars are safer than human-controlled cars. This is already true, and they will only get safer. After 1.8 million miles of driving, Google’s automated cars were in 13 fender-benders, 100% caused by other drivers”||.

    Yes, but I don’t think these were urban areas? Urban environments are hugely complex. We have to bear in mind that these cars don’t see like we do. Many things will throw them. Pot holes, debris on the road, crisp packets, rabbits running across the road and so on and so on. The number of unknowns the environment can throw at the car is unlimited. How do you program the car to distinguish between an empty carrier bag that it can drive through, and a rock? If they brake too suddenly, cars behind driven by people might well crash into the back.

    What happens when ambulances, fire engines etc with the sirens blasting are in the nearby vicinity? What happens with road works? What about the possibility of remote hacking? How are they going to negotiate inclement weather conditions like heavy snow? What of problems such as kids jumping out in front of the cars to see them automatically stop?

    And these cars will be complex. Mechanically complex like other cars, but also complex in terms of all the programming. Suppose a few lines of code cause problems? Happens all the time with computer games and they have to issue patches (which might resolve the original problem, but cause a whole new load of problems). But your life is at stake, you cannot afford to have them programmed incorrectly!

    It will be longer than most people think before autonomous cars have a significant presence on the roads.

  23. Pete A says:


    Autonomous Lexus SUV could not prevent accident that caved in front and rear passenger-side doors, setting off airbags and forcing it to be towed away.

    One of Google’s self-driving cars was involved in one of the worst autonomous vehicle accident yet, when a driver ran a red light and collided with the passenger side door of the modified Lexus SUV.

    See also:

    The essential point of self-driving vehicles is NOT that they sometimes cause accidents; the essential point to assess is the risk-benefit ratio of autonomous vehicles versus the risk-benefit ratio of human-driven vehicles. The passenger aviation industry iteratively reaps the safety benefits provided by automation. Fly-by-wire aircraft are positively beneficial in terms of both their risk-benefit ratio and their cost-benefit ratio.

    Human drivers would cause far fewer accidents if they were regularly subjected to supervision and auditing by advanced driving instructors. Everyone who drives a vehicle while believing that they are a good driver is not a good driver, they are a dangerous driver!

  24. MosBen says:

    JimV, your experience aside, I saw some research referenced in an article about the idea of a basic income which claimed that people report a bump in happiness upon retirement, even if they were unemployed prior to hitting retirement age. Being “retired” is a socially acceptable construct that we’ve created in which it is considered acceptable to not work. It seems awfully likely to me that a lot of the ideas about how people like work are tied pretty closely to social expectations which require that people work to show that they have value; to themselves, their family, and society. If we reach a point in which automation has simply made the demand for human workers permanently lower than the supply I think that we’re going to need both political change to create support for people who don’t have work as well as changing social expectations about the value that we place on work. Maybe the solution to fewer jobs is instead that we all just work less to compensate. Or maybe the government provides a job guarantee paired with a basic income so that people who don’t wish to work will have their basic needs met but anyone who wants a job can get one provided by the government.

  25. SteveA says:

    MosBen: “…changing social expectations about the value that we place on work.”

    I doubt this will ever change. A need for a sense of purpose and usefulness appears to be hard-wired human trait.

    If an individual wasn’t contributing in some way to the survival of their hunter-gatherer tribe/family, then what was the point in helping them through the next famine/injury/illness?

  26. Ian Wardell says:

    ||”I’m retired. Life without some kind of work is not enjoyable, even if you have enough money to live without working”||.

    A lot of work is definitely not enjoyable. A lot of jobs you’re working for someone else from 9 to 5 and you don’t find the work interesting at all, but find it dull and repetitious. The type of work where you occasionally look at your watch and hope 5pm soon comes round. And the weekend! Wishing our lives away. Then feeling gloomy on a Sunday evening as it’s soon going to be the start of another week.

    The thing is we live finite lives. While we’re healthy is it really a good idea to spend most of our daylight hours hoping that the evening and weekend will soon roll round?

    There’s making friends and the general camaraderie with work colleagues. And there’s the issue that we all need to have money! But the point I’d like to make is that there seems to be something fundamentally wrong and unsatisfactory about this whole arrangement. Something fundamentally wrong about the way modern society works. Unsatisfactory and unfulfilling and ultimately dispiriting to our yearning souls.

    Got to say I find it somewhat surprising that many people — if not indeed most people — would find it profoundly boring not being in a full-time job as an employee. That they would have nothing to do all day. Indeed, many people claim that when they were unemployed they were sleeping 12 hours a day and were just depressed.

    I have to say I find this utterly bizarre. So going for walks in the countryside; visiting museums; learning and becoming proficient in some subject and perhaps even becoming an expert; discussions on a variety of subjects on the Internet and elsewhere; reading novels; playing games; exercising; just simply thinking about reality and our place in it and what it all means; laying in a field in the warm bright sunshine in the arms of the one you love. And so on and so on and so on… None of this has any appeal? Peoples’ only aim in life is to sell their labour to an employer? Nothing else in life is worth doing? Wow…

  27. BillyJoe7 says:


    Nice we can agree on some things. 🙂

    I’ve been self employed for 26 years, and it’s the best move I have ever made. I’m the employer, but I know what it’s like to be an employee so I let my employees rule the roost. Which of course means picking the right people. I work about 55 hours a week, but Sunday is my rest day.

    Did I say rest day? No, it’s my activity day and what I imagine life will be like in retirement. My activities mainly involve trial running and cycling. I participated in the “Surfcoast Century” (a 100km trail run) last weekend, and in four weekends from now I’ll be riding “Around The Bay” (a 210km road race). The scenery is magnificent and the cameraderie excellent. On most Sunday mornings I run or ride in the Dandenongs about 5 minutes drive from home.

    And my work is never boring. I can honestly say that I never watch the clock. I have to be reminded that it’s time to go home. I’m one of those people who can switch off from work when I’m at home and switch off from home when I’m at work. Sometimes to the annoyance of my wife. And I can switch off completely on going to bed. I don’t understand insomnia.

    I also read blogs and books but not nearly as much as I would like because of time restraints. Retirement would bring opportunities to indulge in them to my hearts content. I’d like to pick a topic and learn it to death. Become an amateur expert (meaning that I’d pay due deference to actual experts) on a particular subject for which I simply do not get time while working. There’s lots of places I’d like to visit again at a more leisurely pace and places I haven’t had the time to visit at all. I’d like to hone my gardening and culinary skills, and spend lots more time with family and friends.

  28. fbrossea says:

    Some people believe that the things that Ancient Greek Philosophers, the Renaissance, and the industrial revolution had in common is a high density of people tinkering with things. Prior to the industrial revolution, you had a bunch of people making stuff in their garage so it was only a matter of time before someone invented something that would shift things in a big way. Many tech companies follow this lead and allow their employees some “do-whatever-you-want” time with company resources in hope they will happen on an important innovation. It basically ends up following the idea that an innovation could basically come from anywhere so you better hope the person isn’t too busy trying to feed their family and is willing to take risks.

    This, by itself, is likely a flimsy argument to start a universal income program, especially considering it would be hard to picture someone starting a fusion reactor in their garage. It is simply an optimistic view to consider in the wake of increasing unemployment as automation becomes more and more sophisticated.

  29. DanDanNoodles says:

    Ian Wardell:

    Suppose a few lines of code cause problems? Happens all the time with computer games and they have to issue patches (which might resolve the original problem, but cause a whole new load of problems). But your life is at stake, you cannot afford to have them programmed incorrectly!

    This is fundamentally misunderstanding how the programming for autonomous vehicles works. Automotive systems are based on machine learning algorithms, not line-by-line programming like a video game. There aren’t functions with endlessly repeated IF-THEN statements that say “IF THEN “. The programmers do not have to anticipate every single situation.

    ML systems learn how to drive the same way humans learn how to drive: by experience. The difference is that humans have to earn the experience themselves through hours of practice, whereas ML systems can inherit their knowledge directly from other ML systems. Every mile of driving practice by an autonomous vehicle becomes experience for all other vehicles (from the same manufacturer, anyway).

    And here’s the thing — most driving is very routine; it is the unusual situations that require the best driving. The unusual situations, though, are exactly where humans are at their worst, because that is where they have the least experience. But the ML system will have expertise in handling all those situations built in.

    So, you ask, how does a car distinguish between an empty bag it can drive through and a rock? Well, how do you do it? When you see something on the road that looks like a rock, you probably assume it is, until you receive contradictory information, and then adjust your expectations accordingly. ML systems do the same thing.

  30. DanDanNoodles says:

    Oops, my “IF THEN” example got weeded out. I shouldn’t have used angle brackets. It was “IF (situation) THEN (action)”.

  31. MosBen says:

    SteveA, There’s a difference between a sense of purposefulness and what we currently define as work. Right now work is more or less the value of labor that can be extracted from you by an employer in exchange for money, and as a society we place a pretty high social value on a person’s ability to perform work at a high level for high levels of compensation. We also tie a lot of basic needs to this conversion of labor into money so that for most people who aren’t retirement age it is both a social expectation as well as being essential to survival. But while people may always need something to do that they find fulfilling, it may not necessarily mean that this equates to what we currently think of as work. This is especially true if we reach a point where automation eliminates millions, or tens of millions, of jobs which aren’t replaced by new jobs in other fields. If we reach a point at which there simply isn’t enough market demand for employees we’ll either need to drastically change our expectations about what a job looks like (everyone works fewer hours so that nearly everyone can have jobs), or we’ll need to change our expectations about how jobs are created (the government provides a jobs guarantee where anyone who wants a job gets one), or we’ll need to change our expectations about how we value people vis a vis their employment.

    It’s not crazy to think that this can change. The concept of work and jobs have changed over time as economic systems have changed. An economic system that heavily relied on automation for the vast majority of labor will by necessity have a different social relationship to work than ours does.

  32. Bill Openthalt says:

    SteveA —

    A need for a sense of purpose and usefulness appears to be hard-wired human trait.

    Certainly in human males (females of any species are inherently useful). One of the significant reasons for the evolutionary success of humans is that human males have become useful and contribute actively to the success of their progeny, to the point of being indispensable (cf. the singular lack of success of single-parent families, even with massive support from the community). The correlated nearly complete monogamy ensures that almost all males (even the less able ones) have better things to do than destructive competition or the elimination of the offspring of other males.

    This is why using females to perform tasks that can be performed by males is a double-edged sword. On the one hand, it is a fact that many females are more able than many males, and a society will be more productive when it doesn’t “waste” a significant part of the available brain power. On the other hand, males that have no useful role and hence little hope of procreation easily become destructive. This is a more pronounced problem in societies where there is less need for physical strength or protection against enemies and predators. In addition, the devaluation of child-bearing and rearing makes the physically and mentally less able females less attractive to potential mates.

  33. SteveA says:

    MosBen: “There’s a difference between a sense of purposefulness and what we currently define as work. Right now work is more or less the value of labor that can be extracted from you by an employer in exchange for money.”

    In using the term ‘work’ I wasn’t referring solely to paid employment (I would also consider unpaid volunteer work as ‘work’). I was talking about most people’s ingrained need to be considered useful in some way.

    If a person capable of work did nothing but consume, and did nothing to contribute to society in some way, the would be considered a parasite (as opposed to those who, due to age or infirmity, had contributed but were now no longer able to).

    Again, I think this viewpoint is hard-wired, as explained in my previous post. I don’t think that any amount of social engineering (for want of a better term) will put a dent in that.

  34. MosBen says:

    Except that retirement as a concept didn’t exist until relatively recently in human history, and certainly not as an idea that should be accessible to everyone, and yet we’ve incorporated the idea fairly well. Many retirees are not incapable of continuing to work and we don’t judge someone’s worthiness for retirement based on the degree to which they “contributed” in their younger years. Retirement is simply a category where most people accept that people should not have to work unless they want to or desire continued income. Most people simply don’t think of the 66 year old walking around a Target looking for a pair of shorts for his trip to Florida as a parasite, no matter what he was doing when he was 63. And people who reach retirement age and do retire report increased happiness, even if they were unemployed previous to becoming retired. Society simply carved out an exception to our expectations regarding work and everyone just accepted it. I don’t see any reason to think that this was an exception to hard wiring or that some new social pact couldn’t similarly change how we view work.

    Like anything this would take broad social buy in and time to become the social expectation, but I think the burden is on the “hard wired” argument to show that there’s actually something biological happening, rather than just centuries of social expectation.

  35. Lobsterbash says:


    You said:

    “In the end, except for some professional classes of drivers — i.e., police and emergency vehicles — it is impractical to have humans and computers driving together. The temptation to abuse the system is just going to be too strong.”

    You have an interesting point, but there could be designated roads or lanes that are AV-only and human-only. HOV lanes alone could be repurposed for this. I don’t think many would mind having AV-only roads/lanes being loaded up with cameras and monitored closely for violations & abuses.

  36. Gingerbaker says:

    “In the event of unavoidable accident situations, any distinction between individuals based on personal features (age, gender, physical or mental constitution) is impermissible.”

    Can I not even sacrifice myself to preferentially save my loved ones?

  37. Mal says:

    I wonder what affect driverless cars will have on the behaviour of pedestrians. There is probably a minority who are going to become careless and even dangerous when crossing the road because of their perceived immunity. There is also the possibility that some will invent a new variant of the game of chicken (the possibilities for forms of perverse ‘entertainment’ will be manifest).

  38. SteveA says:


    Retirement has been around for centuries; retirement funded by the state is relatively new.

    In humanities formative years we existed in relatively small social groups where your level of contribution would definitely have been noted. Failure to contribute would likely have resulted in ostracism and likely death (because why should we break our backs feeding you when you’ve never done anything to help us?). My contention is that whatever genes contribute to the ‘parasite’ mindset would have been heavily selected against.

    “And people who reach retirement age and do retire report increased happiness, even if they were unemployed previous to becoming retired.”

    I don’t understand what you think this proves. Aside from, ‘who wouldn’t be happy to pick up a pension?’, there are so many potential variables to unpick.

    We both agree that there is a strong expectation that people be productive and useful. You think this ‘construct’ just came about somehow; I think it’s been driven by millennia of evolution.

  39. Sarah says:

    RE: Jobs

    No technology has ever been prevented from coming to market when its time is right. Even GMOs exist in most markets still despite the staggering social difficulties they face.

    There is no stopping the destruction of these jobs. We just have to adapt.

  40. Charon says:

    I think you’re vastly underestimating the difficulty of this technology, Steven. 95% there is immensely, qualitatively different from 100% there. If the car ever has to hand off control to a human driver, it doesn’t work. Either the human has to pay close attention while not doing anything for long periods of time – which is nearly impossible for a human, and removes the whole point of a self-driving car – or the car is going to hand off control to an unprepared person, almost guaranteeing a bad outcome.

    We’re also not addressing the fact that personally owned driverless cars will just make traffic worse. They may be able to avoid some of the human-caused traffic issues, but there will be more cars on the road. People who get work done on buses or trains would be able to drive instead and get work done there. People in cities who hate parking could use their personal cars and then have the car drive off to park itself. Short-haul flights (mentioned above) would be replaced by driving, which is a generally more pleasant experience, but also with larger climate impact (hopefully this latter point is lessened as cars go electric).

    Communally owned cars could help, but that would require an enormous culture shift. Possible, perhaps, but likely? The vast majority of Americans don’t use Uber, Lyft, traditional taxis, or carshares almost ever.

  41. Charon says:

    I’m also interested in the legal questions. Self-driving cars will make mistakes sometimes, and even if they’re more rare than human driver mistakes, they will be different kinds of mistakes. Naively I don’t see our legal system as well equipped to handle adjudicating these sorts of things.

    Are we going to let software developers at car companies determine uniform standards for safe driving conditions? Currently people are allowed to make their own decisions about, for example, when to pull over during foggy conditions. Different people make different calls about how foggy is too foggy.

    Will I, as a cyclist, be liable for the collision if I’m hit by a self-driving car and I hadn’t bought whatever proprietary transmission system that car was looking for?

    There’s already talk of redesigning roads and aspects of cities to accommodate self-driving cars (another sign the tech isn’t there – not a drop-in replacement for humans). Why are we not putting that effort toward reducing car usage instead?

    I foresee autonomous trucks for shipping, at least for the highway portions, but more than that… making rosy predictions is currently unwarranted. This stuff is hard.

  42. Charon says:

    @DanDanNoodles, ML algorithms do not learn the same way humans do. Unsurprising, as they’re not working with the same hardware/wetware. They sometimes “learn” things humans consider bizarre and counterintuitive. Which goes to my point that the mistakes self-driving cars will make will be very different from the mistakes human drivers make. A self-driving car might brake hard for a donut lying in the road, and then plow right through a child on a bike (maybe because the batteries had gone out on the bike’s transmitting beacon). What happens then? There isn’t anyone to send to jail. There’s not even a programmer to blame for a software bug! Society is not equipped for this.

  43. MosBen says:

    SteveA, there have been people who have not needed to work after a certain point for a long time, but the concept that at a certain point literally everyone should be allowed to retire unless they choose not to is relatively recent. As for what the social science surrounding happiness and retirement shows, here a link to an article discussing it:

    I think that it shows that at least a significant portion of the feeling of needing to work comes from a social expectation that a person’s value is derived from work. Creating a social expectation that everyone gets to “retire” at a certain age, thus creating a socially accepted category of non-workers, has changed how people see work and their relationship to it. I don’t see a reason why changing social expectations couldn’t change how people perceive their relationship to work further, especially in the event that obtaining employment becomes literally impossible for some significant number of people absent something like a jobs guarantee.

  44. MosBen says:

    To clarify just a bit. My problem with the “hardwired” argument is that it basically assumes that our feelings about work aren’t learned and are basically impossible to change. Except that within basically one generation, maybe two, of Americans we drastically changed how we thought about retirement. And we have reasonably good evidence that whatever anxiety people have about being unemployed mostly goes away when they assume a social role (retiree) which society accepts as someone who has no expectation of employment to derive their value. That’s not exactly what we would expect with a hardwired need for work. It seems reasonable that people have a hardwired need to feel accepted by their peers or society and that having some kind of activity which they find fulfilling is important. But that’s pretty far removed from what we think of in 2017 as work.

  45. Pete A says:

    Ian, You have reminded me about the insecure design decision made in the Controller Area Network (CAN) Bus. Let’s hope fully-autonomous vehicles never use insecure technology.

  46. Christopher1 says:


    Did you get your nightmare of a 9-5 job from movies, or have you actually ever tried out a 9-5 to see what it’s like? A lot of people think being utterly useless to anyone else is dispiriting. Like there’s no need for you to even exist.

    You mention a lot of things a bum could do instead of work: walk in the countryside, visit museums, become an expert, read, exercise, lay in a field with your lover, etc, etc.

    How many of those things have you done in the last month? In the last year?

    The thing is, it’s hard to do any of those things if you don’t have any money. And other people aren’t gonna pay you just to exist. You have to do something for them to get something from them. C’est la vie. Would you wanna pay an existence tax so everyone else could bum around as much as you do?

  47. Ian Wardell says:


    Most of your questions are an irrelevance. People who are retired will have money. And we don’t exist merely to be useful to others. Once people retire it’s time to enjoy yourself.

    Incidentally, I didn’t disable comments on both of my blogs, and block you on Google plus, just so you can follow me elsewhere on the net.

  48. Christopher1 says:

    I haven’t said anything less relevant than you. The topic was self-driving cars. You’re the one who brought up your work aversion so it’s fair game now.

    I absolutely support retirement. The thing about retirement though is that it is EARNED. You pay your dues first then you get to retire.

    You seem to think we should be born retired. How does that work? How much existence tax should we all pay so you can sit on your rump from cradle to grave? How better off would everyone be if everyone lived like you? No food in the supermarket, nowhere to get gas, no one to rent a flat from cuz people can’t be bothered to work. Not even ONCE in their lives!

    How terrible for you that you can’t censor other blogs. Maybe you would be happier never straying from your safe space?

  49. Drake says:

    Some commenters here assume the majority of self-driving vehicles will be privately owned. That seems unlikely. Not only because autonomous vehicles will likely be very expensive when first introduced, but because there will be a strong disincentive to own and maintain a private autonomous vehicle, regardless of price.

    It’s a bit like owning music recordings: I used to buy CDs and LPs, because they offered convenience (I could listen to what I wanted when I wanted, as long as I owned the recording). I switched to purchasing downloads (albeit at some loss of quality) because the download was somewhat cheaper, and immediate. Now I pay a monthly subscription to a music service, and for the same price as a CD, I have unlimited, immediate access to a library of millions of recordings, with good-enough sound quality, given my consumer-grade stereo.

    Even if I did buy a private autonomous vehicle, hiring it out as a Lyft- or Uber-style taxi during all the hours I wasn’t using it would be a no-brainer.

    An economic argument also suggests new legislation won’t be necessary to limit or ban human-operated vehicles. To succeed at all, it seems like autonomous vehicles will need to be a few orders of magnitude safer (i.e., less culpable for accidents) than human-operated vehicles. Indeed, this already seems to be the case.

    What insurance company would offer affordable rates to human drivers under those conditions? It’s a basic principle of book-making that equal sums of money need to be bet on both sides of the line, regardless of the odds favoring one side or the other. Very quickly, human drivers won’t be able to afford the liability insurance to share the road with autonomous vehicles.

  50. Ian Wardell says:


    I was responding to JimV who is retired. Specifically I was disputing his claim that without work, and even if one has enough money, that there’s not much point to life. Bearing this in mind your posts are simply completely irrelevant. This is in contrast to my own posts.

    We do not know each other. You know nothing about my personal life since I have refused to answer any of your questions. Specifically, you have no idea how I earn a living.

    I think your anger should be directed at the gross inequality that pertains in the USA (I assume, unlike me, that you are a denizen of the USA since you referred to me as a “pussy” in a comment on my blog that I “censored”). Like the fact that the bottom 80% of people in the USA only have ~ 7% of total wealth. Do you think they all don’t work? And the fact that the top 1% have 38% of all wealth with many of them getting their wealth by exploiting the poor.

    I have no idea why you are so obsessed with me and following me around the net sending me unwanted messages. It’s weird behaviour…

  51. chikoppi says:

    The UBI is an interesting premise. There are various propositions and analysis from the economists is both cautious and mixed.

    In one version, the UBI would do away with most of the social assistance programs and tax carve outs (which would drastically reduce federal expenditure) in favor of a cash payout. The amount would be sufficient for basic living costs, but that’s it.

    I don’t know that UBI would be a disincentive to work. People will still want more than they have. What it would do is to provide a cushion or backstop against the encroaching “gig” economy. As jobs become less stable a UBI would allow people to engage in opportunity employment, which would have interesting effects on the labor market and wage distribution. People would be less inclined to take crap jobs for poor wages, which might force less severe disparity between the loading dock worker and the CEO.

    It’s a terribly complex question, which I don’t think can be settled theoretically. Someone somewhere will have to implement it on a limited basis before we can make sound predictions about the practical effects.

  52. Christopher1 says:


    Your problem is you have NO humility. You troll blogs like this just to start arguments because you know it all and everyone else is just daft. If anyone challenges you 100% of the time it’s because they don’t understand your superior arguments. Not once is it because they’re right and you’re wrong.

    Your superiority is one of the worst delusions I’ve seen. Elon Musk is a total idiot, Stephen Hawking is just some third-rate physicist – because YOU know better than them. People listen to them, not to you, because THEY are accomplished and YOU are not. Musk at least is trying to innovate new technologies. What the hell have you ever done? All you ever do is pontificate about how stupid everyone is compared to you – all while never doing a damn thing yourself!

    Sean Carroll actually teaches physics at a state university. What does he have compared to you? A 50 something man who has never had a single job in his whole life! I know because whenever you’re asked if you ever had a job, you refuse to answer. I think you said once you avoid becoming homeless by betting on sports to pay the rent.

    Most rich people have jobs. Elon Musk and Bill Gates have companies to run. So what makes you so special? Why do you get to consume, consume, consume, but never contribute anything to society? Why should we ignore what they say and pay attention to you? They are somebody, you are nobody. They bothered to do something with their lives. You CHOSE not to amount to anything. Don’t complain when no one reads your blog. Why should they? Who are you for people to waste their time on?

    In your delusion education is just a scam because YOU never wanted to learn anything. Kids should just play in the woods all day instead of going to school. They’ll never need to know any of that stuff anyway, because YOU didn’t. It never occurred to you they might want to do something with their lives when they grow up, because YOU didn’t. I think school serves kids who choose to amount to something well. Just because you chose to be a common barfly, don’t take that opportunity away from them!

    I called you a pussy because you won’t allow criticism posted in your safe space blog. It’s part of the song actually –

  53. SteveA says:

    Drake: “Some commenters here assume the majority of self-driving vehicles will be privately owned. That seems unlikely.”

    Agreed. Car leasing is the future.

Leave a Reply