Productivity Gains or Losses via AI Self-Driving Cars: The Inside View

Inside the AI self-driving car, in swivel seats, trying to be productive.

By Lance Eliot, the AI Trends Insider

My daily commute to work takes about 1 ½ hours each way. Here in Southern California, my having a commute that is less than 2 hours each way is considered by some to be a blessing and I ought to relish the “lighter” commute than many of my colleagues. For those of you that live in other parts of the country, the idea of having a commute taking around 2 hours probably sounds crazy. Why would somebody drive that long each day, each way, just for a job?

Up in the Bay Area of Northern California, which I visit each month for work, my colleagues there have at times a similarly lengthy commute, but they tend to be on the BART (Bay Area Rapid Transit) trains during most of the commute time. I at first assumed that they would use that time for a boost in their productivity, being able to do so because they weren’t having to drive a car as they made their commute. Turns out that being inside a train car does not necessarily lend itself to a much greater environment for trying to get things done. You are eyeing other people to make sure you stay safe, and you are contending with the noises in the train car of other people talking (and sometimes singing) and the movement and rattling of the train itself.

For my driving time in Southern California, I often schedule work related phone calls that I figure can be reasonably undertaken while in my car. This can be somewhat dicey though. I’ve found that callers will often realize that I am in my call while in my car, and at times be either insulted by the act or be concerned that they can be overheard. Admittedly, I was taken aback a few weeks ago when I was talking with a work colleague and he was in his car, and all of a sudden I heard his daughter speak-up and ask him a question. I was shocked since he and I were discussing rather sensitive matters at work and I assumed it was he alone in his car.

If I am taking a call during my drive time, I always let someone know if there’s anyone else in the car with me. Turns out that my colleague felt no such need to do so, figuring that what the other person on the phone doesn’t know won’t hurt them. This seems a bit uncivil to me, but anyway, it happens.

Part of my rationale for taking calls while during my driving commute is that it seems to make me more productive. When I am solely focused on driving the car, and perhaps glancing at billboards off the freeway or admiring that sports car up ahead, I often feel like I am not being very productive. Sure, I try to entertain deep thoughts and solve tough problems, doing so in my mind. Many of my colleagues say that the commute driving time is the best time for them to try and figure out problems at work or at home, doing so with the silence in the car and the ability to put their minds to a gnawing issue.

I am sure that you might be concerned about people that take phone calls while driving their cars. We’ve all seen those drivers that seem quite distracted while apparently speaking to a non-visible person, presumably speaking to their phones and not just imagining that someone is in their car. Shouldn’t that person be completely focused on driving instead? The effort to talk on the phone would certainly count as a distracting element and undermine the driving task.

Those that tend to use their phones while in the car, assuming they do so hands-free, would argue that there is no difference between talking with someone on the phone and talking with a passenger in the car. They would contend that if you are going to go berserk about a driver talking on the phone, you should likely go doubly berserk when a driver has a passenger in the car. The basis for being doubly berserk is that the driver might tend to turn their heads and look at the passenger, while with the phone approach the driver can continue to look straight ahead at the roadway. In that theory, the passenger is much more distracting to the driver than would be someone on the phone.

You could try to counter-argue that at least the passenger in the car might be able to aid in the driving task, perhaps serving as a second pair of eyes to be watching the road and maybe alert the driver to something amiss. The phone caller has no idea what’s going on related to the driving of the car. Indeed, a passenger might realize when there are times to shut-up and let the driver concentrate, such as if there is an accident scene up ahead, while the caller on the phone might still be blabbering away and the driver is either too busy to let the caller know that things are hectic or doesn’t want to tip their hand about the difficult driving situation.

I sometimes have several colleagues in my car at once, doing so during my commute or when we are going to a work meeting off-site. This then ups the ante, one might say, in terms of potential distraction for a driver. You have multiple passengers and they are likely to want to engage in dialogue about work or leisure or whatever. Should you also be engaged in those discussions? Well, you are certainly immersed there, sitting in the driver’s seat, so how could you not participate? It would seem odd and perhaps rude to not participate in the discussions. Plus, if you seemed solely concentrating on driving, it makes some people nervous because they figure that if you can’t handle some pleasantry conversations while driving, maybe you aren’t suited to be doing the driving at all.

I’ve often wondered what would happen if there was a conflab of work people in a car and one of them suddenly and unexpectedly told the driver that he had just been fired from his job. Obviously, this might cause the driver to become erratic in their driving. Not a good situation. Hopefully this doesn’t happen very often.

When my kids were young, I would drive them to school before I headed off to work. Sometimes they would try to catch a few extra winks, a bit of extra sleep, and take a kind of short nap on the way to school (it was only about a 15-minute drive). Or, they would try to finish up their homework. This was a questionable practice since they really didn’t have much time to do so, figuring that the time needed to open their workbook, get focused on solving some homework assigned problem, figure out the answer, and right it down, well it was a lot to try and fit it all into a relatively short driving time.

The other problem with studying in the car involved the motion of the vehicle. At times, they would get motion sickness while looking down at their books and papers. This is a rough way to start a day and then end-up at school already feeling nauseous. Furthermore, the movement of the car would make it hard for them to write legibly on their homework papers. If you think that once they started doing their homework on tablet devices or their smartphones that this eliminated the issue of vehicle movement concerns, you’d be wrong. They had as much trouble trying to use the electronic keyboards as they did trying to write with a pen or pencil.

My overarching theme here is about how we use our time while in our cars.

What We Do in Our Cars

When you are driving a car, there is only so much else that you can accomplish other than actually driving the car. When you are a passenger, you can try to accomplish things and have a better chance since you presumably are not required to pay attention to the driving (assuming you trust the driver!). It can be hard though as a passenger to get a lot done since the rocking movement of the car tends to make it difficult to write or read, and you might also suffer from motion sickness while trying to do so. Taking a nap is maybe one handy way to use the time as a passenger, assuming that you are comfortable falling asleep in a moving car (some people aren’t able to do so, either due to anxiety about wanting to be awake in case something goes awry, or due the movement of the car and potential nausea).

Some economists suggest that the time we spend in our cars today is relatively vacant of productivity. The normal passenger car driver is considered negligibly productive since they are focused on the driving task (a role that is perhaps only “productive” in terms of providing transit from point A to point B, but otherwise adds no further value, presumably). Passengers might have some amount of productivity, but it is considered rather low due to the nature of the “work” environment as available in a typical passenger car. All in all, the time we spend in our cars is often considered wasted or under-utilized with respect to being productive.

You can quibble with such a claim. If I am in the car with my work colleagues and we are discussing how to invent the next new mousetrap that will change the world, one could assert that the time in the car was very productive for all of us in the car. If I am in the car with my children while driving them to school, and we discuss the Constitution since they are studying it in their history class at school, I’d say this is productive time for both me and the kids. If I get a phone call from a colleague that is stuck trying to get his computer to function, and I am able to diagnose and offer a solution during the call, it seems to me that I was productive beyond just driving the car.

Overall, we need to try and grapple with the notion of productivity. What is considered as being productive and thus constituting productivity?

Trying to Figure Out Productivity

If you are measuring productivity of a person working on an assembly line in a manufacturing plant, you could in a straightforward manner count how many widgets they are making per hour. Then, if you can make changes to increase how many widgets that person is producing per hour, you’ve presumably increased the productivity of that person. It’s the classic definition of productivity being the equation of output divided by input (for labor productivity, it is customary to use output volume produced as divided by labor input used).

Labor productivity in today’s world is not so easily counted. We have been shifting to a knowledge-based form of workforce. You cannot quite so easily come up with simpleton measures for productivity.

I recall a Service Desk that I was asked to help boost the productivity of the workers or “agents” that were assigned to assist customers. The manager was under pressure to make those workers more productive. When I asked what the measure of productivity was, she indicated that it was the amount of time spent per call. Calls were usually about 10 minutes in length and the average agent was supposed to be doing about 5-6 per hour. The goal to become more productive was to drop the calls to 5 minutes in length and the hope was to double “productivity” to around 10-12 calls per hour.

I pointed out that one quick means to achieve such “productivity” would be to have the agents purposely short their calls. This consists of the agent opting to close-off a call in the shorter time period of 5 minutes, regardless of whether or not the agent had actually solved the customer issue. The customer would likely call again to try and finish the effort, which would be “beneficial” since it would further increase the call volume for the Service Desk. Problem solved!

But, of course, it doesn’t solve the problem and instead creates new problems. The customer is bound to get irked by this trick. The customer will be steamed when they call back. The agent receiving the call will get the brunt of the anger. This will tend to undermine the performance of the agent. It will use up time for the agent to come up-to-speed about whatever was being discussed. And, there will be a number of customers that will opt to not call back at all and take their business elsewhere.

The gist of this example about the Service Desk is that we need to be very careful about how we define productivity.

A narrow definition such as the number of calls an agent takes per hour can be easy to come up with, but it then will lend itself to all sorts of abuses and undermine presumably what you are really wanting to achieve. I got the manager to also add other factors such as customer satisfaction and turn the productivity discussion into a more macroscopic look at the nature of the work being performed, the tasks being asked of the agents, whether tasks could be performed in other manners, whether the calls could be reduced or eliminated by taking other action that would prevent the customers from having to call anyway. And so on.

Are people today being productive in their cars? The answer to the question can differ dramatically depending on how you define productivity.

Applies to AI Self-Driving Cars

You might be wondering what does this have to do with AI self-driving cars?

At the Cybernetic AI Self-Driving Car Institute, we are developing AI software for self-driving cars. One aspect that many are hoping to have happen will be that people will become “more productive” in their cars due to the advent of AI self-driving cars. We think it is a topic worthwhile giving some pointed thought towards.

Allow me to elaborate.

I’d like to first clarify and introduce the notion that there are varying levels of AI self-driving cars. The topmost level is considered Level 5. A Level 5 self-driving car is one that is being driven by the AI and there is no human driver involved. For the design of Level 5 self-driving cars, the auto makers are even removing the gas pedal, brake pedal, and steering wheel, since those are contraptions used by human drivers. The Level 5 self-driving car is not being driven by a human and nor is there an expectation that a human driver will be present in the self-driving car. It’s all on the shoulders of the AI to drive the car.

For self-driving cars less than a Level 5, there must be a human driver present in the car. The human driver is currently considered the responsible party for the acts of the car. The AI and the human driver are co-sharing the driving task. In spite of this co-sharing, the human is supposed to remain fully immersed into the driving task and be ready at all times to perform the driving task. I’ve repeatedly warned about the dangers of this co-sharing arrangement and predicted it will produce many untoward results.

For my overall framework about AI self-driving cars, see my article:

For the levels of self-driving cars, see my article:

For why AI Level 5 self-driving cars are like a moonshot, see my article:

For the dangers of co-sharing the driving task, see my article:

Let’s focus herein on the true Level 5 self-driving car. Much of the comments apply to the less than Level 5 self-driving cars too, but the fully autonomous AI self-driving car will receive the most attention in this discussion.

Here’s the usual steps involved in the AI driving task:

  •         Sensor data collection and interpretation
  •         Sensor fusion
  •         Virtual world model updating
  •         AI action planning
  •         Car controls command issuance

Another key aspect of AI self-driving cars is that they will be driving on our roadways in the midst of human driven cars too. There are some pundits of AI self-driving cars that continually refer to a utopian world in which there are only AI self-driving cars on the public roads. Currently there are about 250+ million conventional cars in the United States alone, and those cars are not going to magically disappear or become true Level 5 AI self-driving cars overnight.

Indeed, the use of human driven cars will last for many years, likely many decades, and the advent of AI self-driving cars will occur while there are still human driven cars on the roads. This is a crucial point since this means that the AI of self-driving cars needs to be able to contend with not just other AI self-driving cars, but also contend with human driven cars. It is easy to envision a simplistic and rather unrealistic world in which all AI self-driving cars are politely interacting with each other and being civil about roadway interactions. That’s not what is going to be happening for the foreseeable future. AI self-driving cars and human driven cars will need to be able to cope with each other.

For my article about the grand convergence that has led us to this moment in time, see:

See my article about the ethical dilemmas facing AI self-driving cars:

For potential regulations about AI self-driving cars, see my article:

For my predictions about AI self-driving cars for the 2020s, 2030s, and 2040s, see my article:

Returning to the topic of productivity while inside a moving car, let’s consider how things might change once we have truly autonomous cars at the Level 5. I do so to allow us to put aside the role of a human driver.

In the case of a true Level 5 self-driving car, the passengers should have no need whatsoever to be involved in or considering the driving of the car. As such, other than perhaps providing commands or conversing with the AI about various elements of the driving, the passengers aren’t going to need to watch the road or look out for pedestrians or take on any other such driving chores.

When I say that the passengers might provide commands or converse with the AI about the driving, this would entail aspects such as telling the AI where you want to go. It might also include that during a driving journey you change your mind about your destination and thusly instruct the AI accordingly. Or, perhaps during the driving journey you get hungry and so ask the AI to take you to the nearest fast food restaurant. And so on. These though are aspects that have no need for the passenger to be engaged in the driving task itself.

For conversing with an AI self-driving car to give driving commands, see my article:

For the socio-behavioral aspects of humans instructing AI self-driving cars, see my article:

For Machine Learning aspects about self-driving cars, see my article:

For more about how humans interact with AI self-driving cars, see my article:

You might wonder whether it is really possible that we will one day be in AI self-driving cars and not have to worry about the driving. In theory, indeed that is what is supposed to ultimately occur.

Right now, it is hard for most people to envision a future of that nature. As such, most of the surveys of people that ask whether they would be willing to be a passenger in a self-driving car are reporting a relatively high frequency of either no’s or that the respondents would do so but then be warily eyeing everything about the driving task (therefore they would in a sense be “engaged” in the driving task, even if there aren’t any driving controls, simply due to their concerns and lack of trust for the AI that’s driving the car).

For the public perception of AI self-driving cars, see my article:

You might be surprised or shocked to know that I actually agree with the respondents that say they would be wary of being a passenger in a fully autonomous AI self-driving car, which makes sense to be wary because if one bases that view on what we have available today, I’d say they are right to be extremely cautious.

I realize you might ask people to think about the future, once we presumably will really have truly autonomous AI self-driving cars in abundance and routinely driving on our roadways. It is almost impossible to today put yourself into those future shoes.

Let’s say some fifty years ago I took a survey and asked people whether they would be willing to carry around a device about the size of a pocketbook and they would be able to use it to call their friends, they would be able to electronically send their friends messages, they would be able to see all kinds of pictures and data from all-around the world on that device, etc.

I dare say that some would assume that such a device would be so expensive that it would be outside the affordability of any ordinary person. They would likely be hesitant to use such a device since they might fear that others could see or read what they are doing.

Imagine the kind of mental leap that would have been needed to envision a world that we have today, consisting of billions of smartphones all around the world and the Internet connections that we have.

I mention all of this because some have suggested that even if we have truly autonomous self-driving cars, some portion of people will be preoccupied with being worried about the driving and it will detract from any kind of “added” productivity that they might gain while inside such a self-driving car. I agree wholeheartedly that if the self-driving car at that juncture is not trustworthy to drive the car, it absolutely makes sense that people would be focused on the driving aspects.

People in The Future

Here’s what I’d like to do herein.

Assume that there will indeed be a period of time for which people will rightfully be cautious and concerned when getting into a truly autonomous AI self-driving car. During that period of time, any productivity will be hampered by the aspect that the people in the car are having to double-check and watch over the AI. It would be like a passenger in a car being driven by a teenage novice driver, and you’d be nervous about every little thing the driver did or did not do.

Next, let’s assume we get past that period of time and exit successfully from the jitters of trusting the AI to drive the car. Assume that we reach a point whereby you trust the AI as much as you would a seasoned human chauffeur that is expert at driving cars. In that case, by-and-large, for most people, you would be able to leave the driving to the AI, doing so without a sense of worry hanging over your head (similar to what most people do when in a limousine or even when in a bus).

Consider a perhaps handy analogy to airplanes. With airplanes, in the early days, most people were rightfully worried about getting onto an airplane. Over time, as airplanes improved and airplane travel became safer, most people have settled down and no longer have much of a concern when they fly. Sure, they still know that there’s a chance of something going awry, and the flight attendants remind you of that potential during the pre-flight instructions, but I’d wager that most people are not especially thinking about the pilot and what the pilot is doing during their flights.

Does this imply that everyone is fine with flying? No, there are of course those that still have hesitations today about flying. The same would likely be true about AI self-driving cars. In that sense, no matter how safe and trustworthy the AI becomes at driving a car, there is still going to be a segment of society that has distrust to the degree that they will be unable to go into an AI self-driving car or will sit on pins during any such journey. I’m going to assume that this segment will be relatively small and it can be considered as a kind of rounding error with respect to people being willing to trust the AI and not be preoccupied with the driving task.

I want to then focus on the time period that involves no particular concern by passengers about the driving of the AI self-driving car. Meanwhile, if you want to argue with me about whether there will ever be such a day, and that maybe we won’t ever get to truly autonomous AI self-driving cars, I’ll offer the thought that yes, I agree it is possible we won’t get there, but anyway for purposes of continuing this discussion, let’s pretend or imagine that it will happen.

For a potential Frankenstein kind of future regarding AI self-driving cars, see my article:

For the use of airplane analogies to AI self-driving cars, both good and bad, see my article:

For the idea of singularity and AI self-driving cars, see my article:

For a Turing test for AI self-driving cars, see my article:

For this discussion, we’ll henceforth assume that the passengers aren’t concerned about the AI driving and those humans are able to devote the minds and attention to just about anything other than the driving.

Right now, any time a conventional car is being driven, it implies that there is a human driver in the car (they are doing the driving), and the human is generally occupied with driving the car. If we eliminate the need for a human driver, this means that anyone inside a car becomes a passenger and no longer a driver.

What might the former human driver that now is a passenger do while inside the AI self-driving car and while the AI self-driving car is on a driving journey?

Notice that I am purposely trying to make a distinction between those people that were already passengers in a conventional car and versus those people in an AI self-driving car that are now passengers and were once drivers. The former drivers are now able to do whatever a passenger can do, meaning no need to deal with the driving of the car. The passengers that were formerly passengers in a human driven car are now able to be passengers in the AI self-driving car, which means they are still passengers as they were before, but now doing so in a car being driven by the AI.

Don’t get hung-up on this. You might say that someone that was formerly a driver might also at times have been a passenger. Yes, of course. I drive myself to work and often for lunch become a passenger in someone else’s car when we drive to a nearby eatery. In that manner, in one day, I am both a driver at one point in time and a passenger in another point in time.

What I am trying to convey is that in the aggregate, we are going to be taking all of those former drivers that spent time driving, and they will in the truly autonomous self-driving car become passengers. This means that all of that time formerly used for driving is no longer being used to drive. This also means that those former drivers can do whatever a passenger can do.

Suppose then that all of those former drivers were to do almost anything “productive” while inside the AI self-driving car. Even if they did something productive for only say one minute, it would mean that if we said before they had zero productivity as a driver, we have now leaped tremendously into their being productive because they now have anything other than zero as their productivity.

Wow, we have mushroomed the productivity immensely, solely by eliminating the human driving task and now having the former human driver be able to provide attention to anything that might be considered productive.

Tricks of Productivity Counting

This is important to note because it shows the trickiness about wanting to predict whether we will be more productive while inside self-driving cars versus conventional cars. You can say it is a slam dunk that we will be, merely by shifting off the driving task and then putting any amount of time toward something considered productive. Note too that it could be that you are productive for just 10 seconds, rather than 1 minute, and you still have boosted productivity, especially if you add it up across the total number of drivers.

Per various stats by the Department of Transportation, there are an estimated 222+ million licensed drivers in the United States and it is claimed that we each spend about 17,600 minutes per year on-the-average driving (we’ll say that’s about 300 hours per year). This suggests that there are 222 million x 300 hours = 66,600 million hours per year in the United States for purposes of driving.

If you assume that driving is a zero-productivity task, this implies that in the United States alone we are “wasting” about 66,600 million hours per year of potential productivity.

If those same drivers opt to become passengers in the same manner as they are being drivers today (number of trips, length of trips, etc.), it suggests that we have a chance of turning the 66,600 million hours into some amount of productive time.

Suppose we are only able to turn 1% of that time into something productive, this means that we still have something on the order of 666 million hours of added productivity per year.

Outstanding! AI self-driving cars that are truly autonomous have a back-of-the-envelope calculation that shows we could boost American productivity by adding over 650 million hours of productive efforts at just a 1% use of their time toward something productive while inside the self-driving car.

If you take a leap of logic and say that those hours are worth at least the value of the national minimum wage (in terms of what people could earn per hour) and use a federal minimum wage of $7.25 per hour, you then have $4,712 millions worth of labor that could be added to the amount of national labor per year. I don’t want to go off the deep-end on this and so let’s just leave it there for the moment (I’m sure my economist colleagues are cringing!).

Of course, we’re only so far considering the productivity of the former drivers. We should also consider the productivity of the former passengers (which, are still passengers, but albeit with a potential for added productivity time if you excise any time previously spent aiding the human driver of the car for undertaking the driving task).

Here’s then what we seem to have:

  •         Former human driving and the hours of those human drivers, which as a driver was considered at a zero productivity, can become a huge boom to productivity by converting those drivers into passengers (or, the driving time into passenger time), and those now passengers can do nearly anything productive, even the littlest bit, and yet still cause leaps in aggregate productivity beyond the former zero.
  •         Former passengers in conventional cars, which already had some amount of productivity since they weren’t doing the driving task, presumably can have at least that same productivity in the AI self-driving cars, and perhaps even more productivity due to no longer being a second pair of eyes for the driver.

Let’s further pursue this matter of the amount of potential productivity involved.

Most pundits predict that AI self-driving cars will be running non-stop and be available whenever you might need a ride. This implies that the availability of using a car, in this case we are saying the truly autonomous Level 5 self-driving cars, allows people to potentially travel more often in a car than they did before. In that case, the amount of time that the former drivers are going to be in a car, and the amount of time that the former passengers were in a car, might all increase (this can be considered a form of “induced” demand).

Thus, whatever we might already consider the added productivity can be boosted even potentially higher because we might have a lot more “passenger” time in the future than we do today.

For my article about induced demand, see:

For my article about non-stop AI self-driving cars, see:

For ridesharing and AI self-driving cars, see my article:

For idealism about AI self-driving cars, see my article:

There is another angle on the productivity question. Will we be able to be more productive in our time outside of being inside a car, due to the advent of AI self-driving cars?

Notice that I’ve only focused on the productivity time while inside an AI self-driving car, but there are some that believe by having AI self-driving cars it will change other aspects of our lives and make us more productive beyond just the act of being inside a car. If you go along with that notion, please add more to the mounting amount of added productivity due to AI self-driving cars.

There are other sides to this coin, which I’ll be addressing in a moment.

For example, you need to be considering that some say we might actually spend less time in the aggregate in our cars once we have AI self-driving cars. If right now you drive your son to baseball practice, you are using your driving time to do so. If you drive your daughter to her activities, once more you are consuming driving time. With the AI self-driving cars of Level 5, you won’t need to go along at all. Thus, we would need to figure out what reduction this has in terms of the conversion of former human drivers that are not going to become passengers per se in circumstances whereby they before needed to drive for some other purpose that didn’t actually need them to be present.

As with most things in life, there will be factors that will add-to and others that will subtract-from these endeavors.

Time is Not a Monolith

In terms of productivity, the discussion herein has suggested that you could have 300 hours per year of potentially productive time handed over to you as a former driver of conventional cars. That seems like a lot of time and could be presumably used for all sorts of nifty things, including perhaps learning a new skill to enhance your existing set of talents. Imagine taking an on-line course that streams into your AI self-driving car. This might not be simply a canned video course but instead a highly interactive class being taught on a MOOC (Massively Open Online Course) basis.

Unfortunately, trying to portray the time as one monolithic chunk of 300 hours is quite misleading. The reality is that it will be maybe 1 hour per day, roughly (I realize that would be 365 hours in a year but give me a break and let’s just say 1 hour per day as an easy approximation).

Worse still, it really isn’t fair to say it is 1 hour per day since the odds are that as a conventional driver you were making around 3 trips per day, and if that continues in the future, it means that each driving journey is only about 20 minutes in length.

The upshot is that the 300 hours is realistically a series of somewhat disconnected 20-minute segments.

I say disconnected because if you drive to work in the morning you’ve done one of the three 20 minute segments, then you work for say 8 hours, you then drive home and get another 20 minute segment, and maybe after getting home for a while then go out to do an errand and get the other 20 minute segment into your 1 hour per day of driving time. These 20-minute segments are not back-to-back. These are segments separated in time and likely your attention and awareness has changed between each segment.

We can even debate whether the full 20-minutes is realistic since there might be time to get settled into the self-driving car and time needed to engage whatever internal system you might use for taking an online class or doing any kind of Skype-like meetings. It could be that the usable time of the 20-minute segments is more akin to 10-15 minutes of actually dedicated and uninterrupted attention.

If you are taking a class while in your car, you would need to have the material divided into rather micro-chapters of 10 to 15 minutes. The YouTube kind of desirable lengths for videos. Even those micro-chapters might need to take a minute or two of the course time to bring you back up-to-speed as to wherever you last left-off.

Another perspective would be to suggest that instead of taking training, you would be performing work tasks of some kind. The work tasks would need to be framed into those 20-minute or less time segments that you are traveling in your car. I had mentioned earlier that I often make phone calls while driving. Those phone calls are usually relatively short and so I probably would get one or two calls during those 20-minutes. Plus, I could do the calls via a Skype-like visual camera since I would no longer need to be focused on the driving task.

I think we should ponder though what people could do for those short segments while in their self-driving cars, and hopefully be something that adds to their productivity. It might be a challenge to do so.

It could be some task that is repeated for each of those 20-minute segments in that they do the same kind of tasks for any of the times they are in the self-driving car, or it could be they do something differently depending on the nature of the timing of the 20-minute segments.

For example, in my case, I do work phone calls during my morning commute and afternoon commute. But if I drive to do an errand after work, I am a lot less likely to be doing a work-related phone call, partially because my fellow workers are also trying to enjoy their evenings too. What might I do with that evening errand’s 20-minute driving segment?

You might also recall that I mentioned my belief that if you are having a call with someone they ought to know that it is private or “public” if someone else is in the car with you. Suppose that the advent of AI self-driving cars pushes us more so toward ridesharing and ride pooling, which means that you might be more likely to have someone else in your self-driving car during your commute than you do now. Will that perhaps inhibit or reduce the ability or desire to make calls while in the self-driving car?

Redesign of Car Interiors

One aspect that seems to be a relatively sure bet is that you might be more likely to do group meetings while inside the self-driving car. The design for future concept cars suggests that the seats in a true Level 5 self-driving car might be swiveled and allow for face-to-face conversations with the occupants. This redesign is aided by the fact that there is no longer a need to put a steering wheel and pedals at the front of the interior and nor dedicate a driver’s seat for driving purposes.

We might also reasonably anticipate that network speeds will be much higher in the age of Level 5 self-driving cars, including the adoption of 5G and then later on 6G. This means that inside of the self-driving car you are likely to be able to carry on group meetings with others that might be scattered all across the globe. Some might be in their offices and others might be in their self-driving cars. I am not saying that the network will be glitch free and nor that it will be the Star Trek kind of reception, but at least it will be smoother and faster than it is today.

For more about the interior changes of cars, see my article:

For my article about how families might interact in an AI self-driving car, see:

For advances in networks such as 5G, see my article:

Will the potential redesign of future cars make it easier to do reading and writing while in a moving car? Recall that I had mentioned the difficulties my children had doing so in a conventional car, while trying to finish up their homework, sometimes at the last minute.

The jury is out still on this question. If you are trying to do conventional reading and writing while in a self-driving car, using pen and paper, I suppose the AI self-driving car is not going to make much of a difference in terms of allowing you to do a better job at reading and writing than a conventional car of today. On the other hand, if the self-driving car has all sorts of touch-sensitive screens mounted throughout the interior, which is being predicted as part of the future redesign, and you are able to also use your voice for auto-transcription into writing, I suppose this new kind of interior arrangement might make it more palatable to do reading and writing (plus, we’ll presumably be doing video watching more than we’ll be doing “reading” of the kind we do today).

Another factor I mentioned earlier was that my children at times would get motion sickness while trying to study in the car. If people are going to try and use their time in Level 5 self-driving car to do studying or otherwise anything other than looking out the windows of the self-driving car, we need to consider whether the likelihood and frequency of motion sickness will rise.

For my prior article about motion sickness in AI self-driving cars, see:

I’d like to take a moment and mention that a reader recently sent me an email about one of my postings and he had brought up two topics, one was his work on the productivity aspects while in self-driving cars and the other was his invention for dealing with motion sickness. His name is Dr. Michael Sivak, Distinguished Research Scientist Emeritus, University of Michigan, and you might find of interest his invention and also his co-authored study on productivity aspects. Here’s a link to his co-authored paper on productivity and self-driving cars:

Another consideration about trying to do something productive while inside a self-driving car brings up an issue that some of the auto makers are already anticipating, namely the dangers of loose objects that could fly throughout the self-driving car and strike someone, which could happen if the AI has to hit the brakes suddenly or make a quick maneuver.

Today’s internal seating tends to prevent loose objects from flying around willy nilly. If we have swivel seats in Level 5 self-driving cars, it means that we are potentially facing each other. This suggests that any loose objects like your smartphone or your bottled water could become an airborne missile that hits another passenger in the face or torso.

The auto makers are struggling with how to protect passengers in general when inside these futuristic redesigned cars. Where will the air bags be? What kind of seat belts will be best? Should loose objects be hooked into some kind of bungee cords to prevent the objects from getting too far from you? I think we can all agree that people of the future will want to bring various loose items into their cars, and we cannot somehow say to them that they need to check their loose gadgets into the trunk.

I had mentioned earlier that my kids sometimes took a quick nap while in our car. For the redesign of future cars, it is believed that the swivel chairs might be shaped to allow you to lean back and sleep, or maybe even be able to remove the swivels and put in place sleeper “seats” when you know for sure that you want to catch some winks.

From a productivity viewpoint, I suppose you might argue that sleeping inside an AI self-driving car is putting us back toward a zero in terms of added productivity. But, look at that point in a different way. If I am able to get some sleep while in my AI self-driving car, maybe it makes me more productive when I get to work. Thus, if you are only counting productivity while inside the self-driving car, it might be unfair because I’ve done something seemingly unproductive inside the self-driving car that made me more productive outside of the self-driving car.

For more about sleeping, see my article:

In spite of the seemingly apparent logic that we will gain productivity by the advent of AI self-driving cars, there are some that suggest we might actually have productivity loses due to AI self-driving cars.

Let’s consider some of the points about the possibility of productivity leakage or losses.

If the interior of an AI self-driving car is ripe with touchscreens and other electronics, perhaps we will use any available time inside the self-driving car for purely entertainment purposes. Maybe we’ll all be watching cat videos and make no effort to better ourselves with the added time that we will have to do something while being ferried by an AI self-driving car.

Perhaps the swivel seats will allow us to have greater comradery with our fellow persons but distract us from doing work.

When you were a driver of a car, you might have been in a more serious mood due to the somber nature of the driving task. This seriousness could have caused you to be work mindful and be thinking work-related thoughts while commuting to the office. In contrast, while in a true AI self-driving car, you could have lost the edge to work and opt to just have fun or be a mental vegetable.

I’d guess that in spite of these potential productivity drains, and though it might lessen some of the productivity gains, it seems hard to imagine that a net effect would be an overall productivity loss. The argument that there will be productivity gains still seems relatively unfazed, though I suppose a bit dented.

Societal Work-Related Changes

When I’ve been mentioning productivity, it has been in the context of work-related productivity.

This will likely mean that if we all start doing work in our cars, and rather than it being something that is happenstance, suppose instead employers come to expect that you will do work while commuting in your car. If that’s the case, it opens up other aspects such as whether you are officially on-the-clock during that time and whether you should be paid for it. In general, there are a slew of laws and regulations related to these kinds of working arrangements.

I’m not suggesting that working in your self-driving car will be a new idea and surprise anyone. Instead, I am merely pointing out that if today in conventional cars we have just a small portion of workers that get work done, imagine the volume and magnitude of work being done while in AI self-driving cars. This would raise tremendously the visibility of working in your cars. I think we could expect new regulations and other concerns that would surface once a large proportion of working society is doing this.

There is another aspect that I lightly touched upon earlier that I’d like to revisit, involving changes in our society due to the advent of AI self-driving cars. Some believe that we might end-up living much further from work as a result of the convenience of AI self-driving cars. Right now, you likely dread having to drive an hour and deal with the high pressures of snarled traffic. Imagine that you were in an AI self-driving car, oblivious to the traffic. And, you were able to work while in your AI self-driving car.

You might decide to live 2 hours or 3 hours away from work, knowing that you can sleep in your AI self-driving car on the way to work, and/or get your work started from your “mobile” office and thus not worry about getting to the office promptly. This again suggests that we might be spending more time in our cars, and maybe a lot more than we do today.

The statistic about drivers in the United States spending 300 hours per year driving is potentially misleading about what will happen in the future. Instead of begrudgingly making that 20-minute commute each-way today, you might welcome with open arms a 60-minute or more commute each-way, allowing you more options of where to live. The impact being that whereas before we were grappling with how to deal with productivity when cut into tiny 20-minute or so segments, it could be that the future will have much longer segments as people actively choose to go longer distances.

We also should consider other kinds of “productivity” besides work specific productivity. Maybe you will be a more productive citizen by using the time in your self-driving car to study up on issues of the day and be better prepared to vote in elections. Maybe you will be able to do more community work, volunteering to aid a non-profit while you are there in your self-driving car and have time to spare. These could have measurable impacts on society as a whole, regardless whether they relate to specifically performing a job that you might have.

For my article about AI self-driving cars as a commodity or not, see:

For my predictions for 2019 about AI self-driving cars, see:

For safety concerns about AI self-driving cars, see my article:

For the potential public shaming of AI self-driving cars, see my article:


With true autonomous AI self-driving cars, it could be the best of times or it could be the worst of times. Maybe we turn toward using our available time in self-driving cars to become a sweatshop and everyone must get onto the treadmill of work the moment they get into their self-driving car. That’s a rather doomsday kind of view. Maybe we are able to use the time in the AI self-driving cars to make some additional money, maybe add more to society by volunteering, and possibly even get to know each other a bit better. I like that scenario a lot more.

When I was driving our car with the kids in it, I relished being with them to take them to school, but I also was trying to watch the road and make sure they got there safely. If I could have been with them and focused on just them, it would have been nice. I dreamed of the day when an AI self-driving car would allow me to do so.

I should end this column there, but I suppose my own kids, now adults, would complain and say that realistically they were probably glad I was distracted so that they would not need to be attentive to my questions about whether they had finished their homework, and that having an AI system drive them to school, doing so without any adult presence, sounds like a dream. Such is life!

Copyright 2018 Dr. Lance Eliot

This content is originally posted on AI Trends.