*The title of this post is an intentional pun, however unfunny the topic.*
Overview:
Autonomous vehicles are a hot topic in today’s technical, economic, and social landscape. Over 30 separate companies are actively pursuing some form of autonomous driving application, some more advanced than others. Regardless, with the consumer market biting at any sort of auto-piloting system available, there’s an argument to be made that fully autonomous vehicles are a plausible inevitability. With the gears of capitalism in full torque, it’s more a question of when, than if.
With that in mind, planning for the transition is an unpleasant task for those who wield political power. Unfortunately, not many are actively considering possible avenues of support for those to be affected by this appending unemployment wave. Now, before I go any further, it’s important to realize that predicting timelines for nascent technologies is almost never accurate. For instance, Google predicted its driverless software to be marketable this year (obviously it’s not), probably fueled by some Kurzweil ideology. Nevertheless, this is something we can see coming, and yet the government seems to favor a more reactive approach. Maybe that’s because they’re slow, or maybe it’s negligence. Either way, if no action is taken proactively, we could see 2% of the U.S. workforce wither into the tech induced graveyard. Luckily, some people are researching this very topic.
In March of this year, just six months ago, the Center for Global Policy Solutions (CGPS) released an exhaustive report on the state of AI driving affairs. There is no way I could paraphrase this magnificent piece, so instead I’ll link you and urge a read of the relatively short executive summary. Report
I agree with most of the report, the only thing I believe lacking were more fleshed out alternatives to the potential jobless drivers. After dwelling on the issue myself for many weeks and reading countless opinion pieces as well as statistics, I have a few of my own recommendations. Some coincide with this report, others are stolen from technoeconomists. (I made that word up) Alas, any new idea I thought I had ended up being something that already existed, but would take major alteration to apply to this issue on a grand scale. Okie dokie, enough preface, lets jump right in.
Scary Statistics:
Whenever anyone starts putting numbers to people, I get really sick. If I took anything away from watching The Big Short, it’s that increased unemployment means more than just human suffering, it means death. The figure in the movie is a bit exaggerated and in reference to the world economy. The real figure for american lives lost due to a 1% raise in unemployment is more around 1,500. You may think, “wow that’s only .0005% of our population”, but I think, “Imagine every person you know dying, times 3”. Needless to say, it’s imperative that as a society we do our very best to ensure a graceful transition to a driver-less nation. (Also, I am only focusing on America for many reasons: 1. The report did too, 2. Driving is way different around the world, 3. I have no idea how half of the other nation’s governments even work 4. Starting “small”)
Big picture statistics provided by CGPS estimate that 4.1 million individuals are employed in driving occupations. This includes: 78% delivery and long haul trucking, 14% bus drivers, and 8% taxi/chauffeurs. Of that 4.1 million, 88% are men and 12% are women. Even more terrifying, 97% don’t have a college degree. This proves to be the most challenging statistic because in order to find employment, most drivers will need re-training. Whether that is getting a degree, or learning a new trade, that will be time and money they may not have.
What will most likely happen before any completely autonomous agents are roaming the roads is autopiloted systems with a driver present. This provides a unique opportunity for drivers to learn/re-train and work at the same time. Depending on the quality of the autopilot system, they may be able to spend most of their time watching lectures of some sort. Leaving the refueling, maintenance, and approach/docking procedures to the driver. More on this later.
The report also mentioned specific states that would be at risk, holding a larger percentage of their workforce in the industry. As much as I think this is helpful for state legislatures to make decisions, I’m not going to highlight it here. Many of my recommendations work for both the federal and state level of governance, but I will speak almost exclusively of the benefits reaped from total federal adoption. Sometimes I think the states are more at odds with DC than they should be, seems like they were designed to work together and often do the opposite.
Possible AI Incapabilities:
After consulting with a few truckers (yes I sought out and talked with those to be affected) they all mentioned a few duties they thought would be difficult to automate, ensuring their job security. These included as already mentioned: refueling, delivery to storefront, maintenance, docking/loading/unloading items. One trucker even mentioned a sort of highway robbery scenario where bandits would stand in the road expecting the system to auto-brake and then bust into the truck to steal all the goodies. As much as I agree with these insider opinions, it’s arguable that these tasks are no more difficult than driving itself.
Hopefully, all these new autonomous vehicles will be electric. And “refueling” could mean one of two things: recharging or a battery swap. Wireless charging has been around for many years, but it’s not awesome. (Apple seems to have flip-flopped with their new X device) Tesla, one of the leaders in the field has been designing autonomous wired charging bots and I’m sure will deliver more robust solutions in the future.
So, the car will be able to recharge itself, what about maintenance? This is by far the more challenging obstacle. Ideally all maintenance is done before each trip at the departure location which could be done by either one person, or maybe someday an automated robot. But, many issues may arise along the road that AI drivers could not possibly anticipate. Emergencies like tire blowouts would seemingly have to have some human interaction.
For delivering/docking/loading/unloading, these would be interesting engineering feats to automate, but I believe a lot of ground could be covered if the processes were redesigned with the autonomous car in mind. Say you’re a keg delivery man; instead of having to get out of the truck, open the door, unload the keg and wheel it to the storefront or whatever, the truck was packed so that each keg came out in order of delivery and could be set in a predetermined location. Of course this is a gross over estimation, but reasonable negotiations could be made per industry I imagine.
Lastly, for the highway robbery case, I would suggest something like a lock-down protocol that called the authorities immediately after recognizing` a threat. Maybe the shell of the truck would become magnetized or electrified. I don’t know, it’s a crazy thought in the first place.
Plausible Inevitability:
As I mentioned earlier, it’s safe to say that this technology will come to fruition. It’s our duty to provide assistance to those who will be affected by this change. Something else to consider is the other industries that may be affected. Many people think car ownership will drop dramatically since we will be able to hail cars, so does that mean all car salespeople will be out of jobs too? This is one aspect of the CGPS report that was mostly neglected. It’s important to think of the secondary industries that can be hurt by this appending transition to autonomous vehicles. The 4.1 million jobs may be a gross underestimation if all the affected industries were accounted for. I could speculate, but I’m not fully aware of the implications and the numerous other secondary industries. Other authors mentioned aftermarket part installers, parking clerks, and insurance dealers. It’s hard to say what the full effect will be.
Needless to say, something needs to be done about the issue. When I thought about possible solutions I came up with 6 major categories of possible support methods, they are:
- “Make New Jobs”
- Outlaw/Regulate the Technology
- Wealth Refactoring
- Workforce Redistribution
- Better Federal Social Support Systems
- Fundamentally Redefining Work
I will discuss each in detail in the following section.
Alternatives:
“Make New Jobs”
I put this in quotes because it’s something many economists preach as the best solution to tech induced unemployment. In the past decades this has certainly been true, and will somewhat be in the near future. But the important thing to realize is the rate of creation. Currently job creation has stagnated. And after witnessing the mass unemployment wave in 2008, it should be obvious that it’s not always easy to get a new job. Most new jobs are in small technology based firms or self-employed entrepreneurs. The report mentions the necessity for entrepreneur education in secondary school. This, along with many other things, would be important to teach to our youth. Maybe some day I will write a post on the necessary changes to America’s education system, but for now I don’t think it will be enough.
New jobs nowadays most often require a college education, which is something these drivers don’t have. What I think this should mean is expanding the workforce in a few select industries, construction being one of them. It will be a while until we automate the building and repairing of our domestic infrastructure, ensuring job security. Of course, this still requires a targeted skill set that the drivers may not have. But instead of relying on new jobs to spring out of nowhere, we should target markets that are currently experiencing worker deficits. More on this later.
Outlaw/Regulate the Technology
An obvious way to ensure these individuals won’t lose their jobs is to completely outlaw the technology. Not only is this morally repugnant, this is counterproductive to any policy ideology. When visiting Oregon I was disgusted to find out that a law passed in 1951 prohibited me from pumping my own gas. This law has two main justifications, safety and jobs. That’s bullshit. Not only has nozzle/pump technology been greatly advanced in the past 60 years, but those jobs hurt the economy more than help. Costing citizens extra pennies at the pump to pay for a service they can 100% do themselves.
In general, a law should never be passed to protect jobs, only to create more purposeful ones. So, outlawing the technology to protect these jobs is bad practice, but regulation is a necessity. Similar to how there are pump regulations ensuring safety, these AI systems need to be strictly regulated in order to insure public safety. Extensive field tests are necessary, and an independent committee should be formed to create them. Manufactures and the public should be consulted to find an agreeable form of autonomous driving regulation. Let’s make sure this is as bi-partisan and unbiased as possible. After all, vehicles are the most deadly weapon we’ve ever made based on the sheer numbers.
So yeah, I know government is already way to big, but we need another agency. ADC, Autonomous Driving Committee, or something, to police these up and coming tech giants, create comprehensive regulation and administer the tests and data collection of these systems. Luckily, it’s on the list of things to vote on, but missed the summer recess. Hopefully they hash things out when they return.
Wealth Refactoring
This without a doubt will need to happen. Now I’m not asking to abolish capitalism for socialism, but think about the increased profit to be expected by these trucking/taxi companies. Trucking companies spend around 30% paying their employees, that would mean an immediate 30% profit boost. Not to mention the increased hours of operation allowed by the technology, boosting the total profits even more. Companies like Uber take around 20% from their drivers. With autonomous agents, that would be a 500% profit increase. Not only should these companies pay for necessary infrastructure updates (if any), they should expect to be taxed more. Those returns should go directly to those programs dedicated to helping the recently unemployed. I’m not suggesting gutting their profits to match pre-AI figures, but a fair percentage should be agreed upon.
Workforce Redistribution
As mentioned in the Outlaw/Regulate section this is a must. The important questions are how, where, and when. (who, what, why is obvious at this point) I haven’t thought of all the possible industries, nor all possible avenues of opportunity. These are merely my thoughts and ideas after reading many books on the subject and consulting with some of the data.
Instead of looking at the workers first, lets look at the job shortages in america. Its important to look at job necessity trends as well as current need to properly estimate the lacking job markets in 20-30 years. But most economists agree across pretty much all developed nations that the three things we will need more workers in are construction, engineering/manufacturing, and caretakers. The workers should have all the choice in what they want to do, but the government should entice these industries in some form. Not only that, but we especially don’t want workers going into programs that will waste their time and money getting a degree that doesn’t award them an income.
Construction workers are needed because a much of our infrastructure is decaying. Engineers and manufacturers are needed because we need to keep developing solutions to scientific and technology based problems. And care takers because of all the soon to be very old people. The government can sponsor work placement programs that host educational content and guidance on the importance and usefulness of these careers.
Next up is to think of how to do this, Jerry Kaplan, an optimistic economist, (yeah they exist) had the idea of a job mortgage. Companies would agree to hire an individual if they passed some educational qualifier, and the individual would take out a loan against their future paycheck. Thiscould be a financial tool for the individual and a motivator for qualification programs to create a competitive learning environment. The companies get tax write-offs, the banks get interest, the workers get a new job: somehow everybody gets paid.
The report mentioned an increase in apprenticeship, which wouldn’t be too bad. But what I think their best suggestion was, and this is something quite controversial, lowering the cost of a college education. The government pays more for not only these workers who want to gain a degree, but young individuals too. After all, those are the groups facing harshest unemployment. Fuck it, even if they study music, at least they are able to contribute to society again. It’s of dire importance to find another ‘thing’ to do, even if it is not going to give them income (I have a solution to this later). Because the data for middle aged unemployment is already getting scary.
Last is when, and this question varies the most due to the proposed timeline being so fuzzy. In my opinion, I believe that there will be a period of regulation and testing amounting to at least 5 years before the government deems them safe enough to be truly unmanned. This may have already started. Even after that, the companies may stick with a driver is they choose to do so. Not for long I bet, maybe 15 more years until human driving is outlawed entirely (in America). So after a truly revolutionary and complete level 5 autonomous application is developed, 5 years with a mandated driver, then 15 years after humans are not allowed. I truly expect a 20 year cycle from driver to driverless. In the 5 years of a mandated diver, the workers should spend time thinking about potential exit strategies, and maybe even start introductory courses in the subject of their choosing. Many things can be learned online today, and if the driver was only needed for a fraction of the time, they could listen to lectures/watch videos when they could spare the opportunity. This would greatly soften the blow to the transition. Then hopefully immediately after they get laid off by the company, they can start pursuing their new career path. Or maybe the companies work with the employees to transition immediately on a scheduled date.
Better Federal Social Support Systems
The report mentioned the need for automatic enrollment in services like unemployment insurance and medicaid. As of now, these programs also need a massive funding increase. Today in America, only 25% of those who applied and qualified for UI were given it, pretty messed up right? We pay so much in taxes, yet so much of it goes to useless purposes, and year after year budgets get cut for the most important services.
I propose the need to change dramatically the way we provide for the old, poor, sick and defeated. First off, nobody should have to live in squalor. Nobody should be denied a service they qualify for and need. Americans should have a standard of living higher than any other nation. How do we do this? We re-imagine social support provided by the government.
Social security is probably the greatest thing that America has ever implemented, in fact it’s also the most reliable. The fact that it is still helping people 80 after its inception is frankly remarkable. But it’s failing. The baby boomers are growing older, and birth rates have been at all time lows for decades. This means there are no longer enough workers to pay for all the recipients, and people are growing older than ever before. My grandma suggests taking more from workers to compensate, I suggest we do something completely different. The first thing you can do is remove the income cap, which always sounds downright illegal. The second is progressive basic income (PBI).
Before you puke on your screen, and cast me aside like the socialist I am, heard me out. Progressive basic income is very similar to social security, and would not replace it, but instead be supplementary to the service. Slightly different than universal basic income, which grants every individual the same monthly benefits, progressive basic income is a calculated needs based government supplement. It would never be enough to live off of exclusively, however nor was social security but over 33% of its recipients make it so. When it comes to the very poor and maybe homeless, rich people say “get a job”, then I say, how can they do that without X. PBI could give these people the opportunity to get that X, whether it transportation, clothes, or food. The reason I favor progressive vs. universal is because many people don’t need this, and their share of UBI could go towards the PBI share of the very unfortunate. PBI, automatic UI, and modified Medicare/Medicaid would cushion any job loss, not just those lost due to trucking.
Fundamentally Redefining Work
This section is reserved for the last because it is the most radical and least time sensitive. I’ve many a time imagined a world where less than the majority of people work a conventional job as we know it today. Would half of our population be on the streets? No, because at this point we would have enacted many of the above policies and probably more. Nevertheless people still need something to do, and should be compensated in some form for whatever it is. So what could a solution for this be?
Community service crowd-sourced income. Sites like JustServe and GiveGab sort of offer the services I’m talking about. My ideology is that the government would run a service like this in which every non-profit and interested person could make an account. Normally the government is crap at hosting online services like this, but I can imagine a day where they hire more software developers to create and maintain a platform like this. Or maybe they’ll just contract it out. Either way, this would be a way for people to do things on a day to day basis, get them involved in their community and provide small monetary sustenance. Where’s the money come from you ask? I had the same question when looking at these already existing services. The service providers aren’t themselves non-profits right? It mainly comes from organization membership fees, which in turn means it comes from government stimuli to the non-profit, and donations. If the government took out the middle man, they could charge organizations less, and also pay the workers a bit.
I know it’s odd, community service for money, but you have to realize the difference between volunteer and community service. One of which you usually get a reward other than that “good feeling”. It may be far-fetched, but this would pay huge dividends for public good and increased quality of life. Communities would be safer and more happy. The logistics would need to be discussed and argued about, but the main idea seems pleasant to me. Say its 2040, I’m unemployed and receiving welfare checks that pay for my modest lifestyle. But I want to get more involved with the people around me and make some extra money because I want to start a new hobby. If a popular service was in place, I could wake up in the morning, sign up for an event, work a shift and get a slight increase on my next month’s check. An app would track this for me, and have all the tools for my welfare need. I say slight increase thinking a few dollars, but it would all depend on the job. Organizations could set all sorts of settings on the app for who could do the work and for how much, and the service would automatically grant or deny the job opportunity based on the percentage of total budget allotted to the organization over a specific period. Remember, this is when more than a majority of capable workers are unemployed.
Conclusion:
I may not have covered all my bases, but I hope I’ve at least convinced you of the importance of this problem. The need for solutions is there, and time is always ticking. We as a society need to recognize the issues involved with transitioning to a driverless nation and demand that our governments act proactively to solve them. From regulation to worker redistribution, this issue needs a compilation of efforts to ensure gracefulness. If done correctly, the building blocks could be set in place for other potential AI induced industry disruptions. Lets try not to have another recession, let’s take care of our people. Comment below your curiosities and arguments.