When the Machines Come (2.0)

By Alexander Greco

Artificial Intelligence is coming. In fact, it’s already here—it’s already been here, in its early stages.

We saw automation occur in small, almost laughable ways, like switchboard operators and bowling pin re-setters. Then automation partially replaced bankers with ATM’s, and factory and manufacturing jobs. Bridge toll collection is becoming automated. Drones are becoming like automated fighter pilots. Soon we’ll have automated cars, which means trucking and taxi driving will be automated.

There’s been a narrative in America the last few years that immigrants and foreigners are taking all our jobs. This is partially true. Sure, a whole lot of work gets outsourced to and from America to people across the world, but this doesn’t account for the whole pie.

A vast number of jobs has already been lost to automation, and it seems highly probable more will soon be replaced.

According to the Ball State University, Center for Business and Economic Research, over 5 million manufacturing jobs have been lost due to increasing automation in factories.[1] This isn’t going to stop, it’s only going to increase.

Artificial Intelligence is being developed at a rapid rate. Various forms of Artificial Intelligence have already beat the world’s best Poker[2], Go[3], and Dota 2[4] players. In addition, AI has begun to:

  • Read handwriting
  • Classify images
  • Recognize objects
  • Understand speech
  • Explain itself (sometimes referred to as Explainability)
  • Self correct

This names just a few groundbreaking capabilities AI has, which, until now, only humans could do with any high-level proficiency.[5]

Let’s synthesize a few of these things and see what we get. Let’s say you have a machine that can understand human speech, it is able to recognize and classify objects, and it can communicate with its “supervisor” to tell them why they did a certain thing. You tell it to sort all the freight in a retail store, and it does so faster than the hardest working employees at the store.

  • It moves with near-perfect efficiency
  • It doesn’t get bored (or bitter, resentful, or passive aggressive)
  • It doesn’t get tired
  • It doesn’t need any bathroom breaks or lunches

In one day, it does the same amount of work as three employees. If it makes any mistakes, the new, more sophisticated forms of AI will be able to correct their own behavior, and it won’t forget the correction the next day (since AI won’t have the same memory problems humans do). People can’t compete with this. Automation is going to take over a huge portion of jobs in America, and it’s going to happen much sooner than most people are prepared for.

Here are the top 15 jobs in America[6] [7], as well as the likelihood that AI will replace that job in the next 10-20 years[8]:

Occupation Likelihood of Automation
Retail Salesperson 92%
Cashier 97%
Office Clerk 96%
Food Prep and Service 87%
Registered Nurse 1%
Waiters and Waitresses 94%
Customer Service Reps 55%
Material Movers 85%
Janitors 66%
Warehouse Stock Clerk & Order Filler 64%
Secretary 96%
Bookkeeping, Accounting & Auditing Clerks 94%
General Managers 16%
Tractor & Trailer Drivers 79%
Elementary School Teacher 0%

These top 15 jobs are comprised of millions of people, an enormous proportion of the nation’s workforce. Millions of people. There are nearly 5 million people working as retails salespersons, 3.4 million working as cashiers, and around 3.1 million people working as office clerks. The top three occupations are held by over 11 million people, and those are three of the most likely to be lost due to automation. The only people who are at all safe on this list are registered nurses, general managers, and elementary school teachers.

And this isn’t anywhere near a comprehensive list of jobs that might be replaced with AI. The following list isn’t even fully comprehensive either, yet it illustrates the expansiveness of the AI Revolution:[9]

  • Factory Worker
  • Chef
  • Surgeon
  • Security Guard
  • Pharmacist
  • Food Delivery
  • Journalist
  • Soldier
  • Receptionist
  • Telephone Salesperson
  • Construction Worker
  • Tour Guide
  • Mixologist/Bar Tender
  • Compensation and Benefits Manager
  • Advertising Salesperson
  • Low-level Legal Work
  • Couriers
  • Proofreading
  • Computer Support
  • Market Research Analyst
  • Data Entry Clerk
  • Taxi Driver
  • Insurance Underwriter
  • Claims Representative
  • Bank Teller/Rep
  • Financial Analyst
  • Inventory Manager
  • Inventory Stockers

The list goes on, but what can be learned from the list?

AI will primarily take over jobs that involve:

  • Repetition or Routine
  • Algorithmic or Methodical Tasks
  • Data Entry
  • Data Analysis
  • High Levels of Vigilance or Monitoring

AI will primarily take over jobs with predetermined routines or procedures, or jobs that require predetermined responses to events. In a factory setting, an automated machine will be programmed to repeatedly perform the same task over and over again. In a bar or restaurant setting, an order will be placed, and the machine will respond by going through whatever pre-programmed procedure it has to make that order.

As a security guard, a robot will monitor its surrounding, and if there is a set of stimulus that the robot perceives to be some sort of threat, it will set off an alarm or investigate. With self-driving taxi cars (which will become obsolete if you own your own self-driving car), you tell the car where you want to go, and the car uses GPS to determine the route it goes. With the variety of software we have now, there are already some automated versions of proofreading. Even Microsoft Word has a simple form of proofreading that alerts you if something is misspelled or grammatically incorrect. We even have software that can tell if a text has been plagiarized.

This means two things.

One. It means there will be a whole lot of shitty jobs that people don’t actually like (even a lot of those jobs where people tell themselves, “You know, I actually like my job.”) that will be taken over by machines. It means that many people will be liberated from the drudgery of their 9-5. It means so much of the work that people can’t stand will be automated by machines that are much better at doing these jobs than we are.

Two. It means there will be a whole lot of people out of work. I don’t just mean a whole lot, I mean a WHOLE lot of people. MILLIONS.

Well, you might say, they can just go find another job.

All the jobs are taken, or the jobs that will be left will likely require some form of higher education, certification or specialized skills. Here are the jobs that will remain:

  • Executive or Leadership/Management Roles
  • Educational Jobs, Like Teaching
  • Human Resource Jobs
  • Engineering Jobs
  • Financial or Legal Advising Jobs
  • Medical Jobs (some, a lot are going to be taken over)
  • Creative and Entertainment Jobs (which aren’t known for their widespread, lucrative outcomes)
  • And a Handful of Vocational or Trade-School Jobs (though not all)

This is quite a list, and each of these fields are rather broad, but the jobs that remain in these fields will require some form of higher education or training. In addition, there will be jobs that remain in the Computer Science fields, and some jobs that will be created by AI, but these almost exclusively require expansive knowledge in Programming, Software Design, Computer Engineering, and so forth.

So, to summarize this all, many jobs will be taken over by AI. Most if not all of these jobs are jobs that people don’t particularly enjoy. However, whether or not they enjoy these jobs, the people occupying these jobs need them to pay their bills. If they want to get another job once theirs is lost, many will have to go back to school (which means either draining their savings, or going into debt for student loans), or receive training in an occupation they have little to no experience in.

Now we come to the question, “What Will Happen When the Machines Come?”

Having AI will be a huge economic advantage to any company using them. Because of the productivity and efficiency of AI machinery, automation will rapidly pay for itself, and churn out more products, more services and more completed tasks than any human could ever have done. This means there will be more goods and services provided to society, presumably at a cheaper price.

This is an important note. In theory, automating the production of goods and services will make them cheaper.

Having AI also means that many people will be out of work. Some of those people may be able to find work, some of those people may not.

Now, to begin understanding what may happen as a result of this, we should look at two previous parallels in recent history:

  1. The Luddite Movement during the Industrial Revolution
  2. The Great Depression

The Luddites were a faction of English textile workers who protested the automation of textile factories by destroying textile machinery[10] [11]. Why? Because, as horrible as it was to work in one of these factories, that’s how they survived. They provided food and shelter for their family by working long hours in harsh conditions. When their jobs became automated, it meant that all their skills and work experience would be a waste, and that they would no longer have a means to provide for their family.

Imagine something like this happening on a much grander scale.

What will a truck driver do when their $40,000-80,000 salary disappears, and is replaced by a self-driving freight truck? What will that trucker do if they’re not certified for any other well-paying job, they no longer have the funds to get any new form of education, and all the entry-level jobs that don’t require training are gone? What will millions of truckers do when their salary disappears because a self-driving truck took their job, they’re unable to receive training in another vocation, and most entry-level jobs are automated?

Now, what will happen when millions (MILLIONS) of retail workers, taxi drivers, office clerks,  factory workers, and so on face similar predicaments?

The next parallel, The Great Depression, in a nutshell (for anyone unfamiliar with or rusty on the subject), happened in the late 1920’s and lasted until the late 1930’s. It began with a huge crash in the stock market, where the worldwide GDP dropped around 15%.[12] During this time, unemployment peaked in the United States at around 25%[13] (meaning 1 in 4 American citizens didn’t have a job), and by the end of the Great Depression, an estimated 50% of Americans lived below the poverty line (this meant that 50% of Americans lived on less than ~50 cents per day)[14].

Why does this parallel matter for when AI comes?

With AI, prices of goods and services will drop drastically (in theory), but so will the wages of everyday people. In many cases, the wages of everyday people will drop to $0. It is possible we will see rates of unemployment and poverty increase drastically as AI takes over more and more jobs, and this doesn’t even account for the potential economic bubble we may be in right now (which I may write about in the future).[15] [16] [17]

Even the overall cost of living, for everyone in America, should drop substantially once AI becomes more and more prevalent in different industries, but is that much consolation for the mass of people who might become unemployed by this?

It is highly possible we will see a mix of the Luddite Movement from the Industrial Revolution, and the economic collapse of the Great Depression. Massive amounts of products will be created, but massive amounts of workers will be displaced.

But, life in America will be improved drastically with more and more automation.

This is a crazy situation. It’s a technological revolution, and an economic conundrum we’re about to face.

So, what do we do?

Assuming we don’t want massive riots, or an even worse Great Depression, we have to figure out a way to survive this coming revolution.

There’s a handful of popular, generalized perspectives on what may happen, and how we should prepare:

  • AI is not a threat, nor will it be a threat in the near future OR if AI does begin replacing jobs, it will not replace as many as people project, and it will replace jobs slowly
  • AI is one of the greatest existential threats humans will have ever faced, and we should be highly cautious and pro-active about AI
  • AI will be massively beneficial to the human race, and people across the world will see an increased standard of living and well-being
  • The Libertarian Approach: If AI begins replacing jobs, and if the economy shifts, let the markets correct themselves.
  • The Social Free Market Approach: If AI begins replacing jobs, we should redistribute the wealth accumulated by Automation in order to offset the loss of jobs, and keep the economy functioning
  • The Regulatory Approach: The use of AI and the potential detriments and benefits should be monitored and regulated by the government.

Personally, I am rather optimistic about AI, but I am also fully aware the possible risks we face with AI (not only with automation, but with the potential supra-human intelligence of AGI). I teeter between, “This will be the greatest things humans have ever created”, and “This is an existential crisis for humanity.”

As far as economic policies, I am torn between laissez faire economics (we’ll call it the Libertarian Option), social redistribution (the Social Free-Market Option), and regulation (the Authoritarian Option). I see merit in all of them, but I also see dangers in all of them.

  • The Libertarian Option
    • The economy will self-adjust, so that as people lose jobs and cannot make money, the prices of goods and services will naturally deflate.
    • However, there is the potential of a “Winner-Takes-All” outcome, where the few companies who successfully utilized the most advanced forms of AI will quickly out-compete all other companies.
    • There is a possibility of a cultural and individualized economy (something like a decentralized economy), but there is also potential for oligarchical rule by tech companies.
  • The Social Free Market Option
    • To offset the economic changes caused by AI, and to offset national job loss, citizens should be given a Universal Basic Income (UBI) to at least pay for food, gas, insurance and so forth.
    • The problem with this is that it tampers with the natural flow of economics (and thus could negate any positive re-balances of an economic low), and it also runs dangerously close to Socialist and Communist politics/economics (though it is technically neither).
    • This Option may reduce the suffering of people that have lost their jobs, but it may also negatively influence the economy in ways that will lead to future suffering and economic failure.
  • The Authoritarian Option (or, the Current-Economics Option)
    • To limit the potential negatives of automation, the government will regulate the use of AI. This will also regulate the use of AI by third-party users, government officials (in theory), and any potential “bad actors” who might use AI.
    • However, any sort of regulation (especially when regulation borders on Authoritarian control) could result in corruption, difficulties for small and large businesses to utilize or develop new technologies, and could lead to social/political tensions.
    • This option may mitigate the reckless use of AI, but it might also centralize power even more than laissez faire Capitalism.

Out of these, I want to focus on the Libertarian Option and the Social Free Market Option—although, despite the fact I am personally not a fan of the “Authoritarian” Option, as my name implies, it is certainly not off the table in political debate.

The Libertarian Option revolves on the idea that the economy will deflate rapidly, so that even people with very little money will be able to pay for goods and services, and that people will become more self-reliant and financially independent in the absence of our inflated economy. In the Libertarian sense, most people are quite adaptable and will find ways to make it in the new economy.

Now, the Social Free-Market solution would utilize UBI (Universal Basic Income). The running idea with Universal Basic Income is that every person old enough to work will be given a $1000 dividend per month to live off of. The $1000 dollars a month is just enough for people to buy groceries, pay for gas, and maybe pay rent and utilities (maybe).

Part of this money would come from the massive savings that automation will bring, cutting spending from things like Welfare, and from VATs (Value Added Taxes), which will be collected from companies utilizing automation. This money will provide a financial safety net for people, in addition to stimulating the economy.

If half the country is living off of a few dollars a day, how can there be any economic growth? If half the country has at least $1000 dollars to spend a month, most of that money will go back into the economy, and people will be able to at least survive.

So, out of the two of these, which is a better option?

I’ve spent a lot of time thinking about this, and I think the best solution is a mix of the two.

We are currently at the peak of human achievement. The things we’ve created in the modern age might’ve been seen as tools of the gods in ancient civilizations. The ability to talk to people across the world. The ability to pull a plastic rectangle from your pocket, and have all the world’s knowledge at your fingertips. The ability to write a blog post, upload it to this global net of information, and display it to millions of other people.

We are currently at the peak of human achievement, and we’ve never had more opportunities to be the masters of our own lives than ever before.

It’s never been easier to survive and flourish than ever before.

Just imagine the grassy meadow on the other side of the AI Revolution. That warm, sunny meadow where we’re all self-employed, or we managed to push ourselves into getting a job we’re proud of, or we’ve learned to live with less and stop putting so much emphasis on material wealth. If people want to rise to the top, they’re free to. If people don’t mind living a simpler life, at least they’ll be fed, and their families will be fed.

But, if we want to get to those grassy meadows on the other side, we have to get over this coming hill. If we can figure out a path forward once Automation comes, and we can hold this ship together while the economy (not just the American economy, the global economy) realigns itself, then we can make it to that grassy meadow.

Let’s say AI comes, and it displaces half of all workers.

  1. The Economy deflates:
    1. Wages drop (for everyone, low, middle and upper class)
    1. Prices drop
    1. Cost of living drops
  2. Everyone is given a $1,000 dividend to make ends meet.
  3. Since the economy is deflated, that $1,000 goes further than where it goes now
  4. But, everyone is still out of work, and many people (particularly families with children) are still struggling to get by
  5. So people learn how to be self-reliant, and people learn how to adapt. The culture shifts to a culture of making money on your own terms, by your own means, or the culture shifts into a culture of mastery, where more individuals receive the training or education to enter a more skilled workforce.

Not everyone will be able to adapt, but maybe we can at least help them survive. Not everyone will be successfully self-reliant, but they can survive. Not everyone will have a great lot in life, but at least they won’t be scared about their children starving.

And if the society can at least survive, then there will be people who learn to live happy and successful lives in this environment. There will be able who discover a way to make a living doing something they actually enjoy. There will be people who learn to live healthy, fulfilled lives, even without a $1,000 dollar dividend.

We can become a society that learns to live without the excesses and the inflation we have today. Despite the fact that multi-billion-dollar companies will have access to armies of robot-workers, I think we’ll see a positive change in income inequality. All the bullshit jobs we work right now, all the money we scramble to receive, and all the rising prices we have to put up with, the value of these things will drastically drop.

And I think the value of these things will drop not only economically, but culturally as well. How many shitty jobs are there? How many shitty jobs are there where you’re forced to smile and say, “I love it here”? Is there anything better you’d like to do with your 40+ hour work week?

My hope is that we’ll see just how unnecessary so much of our stress and struggles are, that maybe we’ll stop caring so much about money, and learn to live with different value systems.

So, when the machines come, I’m hoping we see a drastic shift in the way we live our lives, and the way we view our lives. Maybe we’ll see how worthless a lot of these jobs are (tell me, will you really miss it?) if they can be replaced so easily by machines. Maybe we’ll remember all the things we wanted to do with our lives before we stumbled into this mad struggle. Maybe we’ll learn to live without big corporations telling us how much we’re worth, and without a government holding our hand and spanking us when we’re bad.

Maybe, when the machines come, we’ll learn how to be human again. Maybe we’ll learn how to live independently again, and come together as communities more. Maybe we’ll learn how to be happy again.

[1] Michael J. Hicks and Srikant Devaraj. 2015 & 2017. “The Myth and Reality of Manufacturing in America”. Ball State University Center for Business and Economic Research.

[2] www.cmu.edu/news/stories/archives/2017/january/AI-beats-poker-pros.html

[3] https://en.m.wikipedia.org/wiki/AlphaGo

[4] https://openai.com/five/

[5] https://en.m.wikipedia.org/wiki/Artificial_intelligence

[6] Bureau of Labor Statistics, U.S. Department of Labor, The Economics Daily, Retail salespersons and cashiers were occupations with highest employment in May 2015: https://www.bls.gov/opub/ted/2016/retail-salespersons-and-cashiers-were-occupations-with-highest-employment-in-may-2015.htm

[7] https://www.google.com/amp/s/www.aol.com/amp/2010/11/11/americas-most-popular-jobs/

[8] https://willrobotstakemyjob.com

[9] https://www.google.com/amp/s/www.telegraph.co.uk/news/2017/09/27/jobs-risk-automation-according-oxford-university-one/amp/

[10] https://www.smithsonianmag.com/history/what-the-luddites-really-fought-against-264412/?no-ist=&page=2

[11] https://www.history.com/news/who-were-the-luddites

[12] Garraty, Great Depression 1986

[13] ISBN 0-07-319397-6, Frank, Robert H.; Bernanke, Ben S., Principles of Macroeconomics (3rd ed.). Boston: Mcgraw-hill/Irwin, 2007

[14] https://history.stackexchange.com/questions/12264/during-the-great-depression-roughly-what-proportion-of-americans-lived-below-th

[15] https://www.lynalden.com/economic-bubbles/

[16] https://www.google.com/amp/s/www.forbes.com/sites/jessecolombo/2018/09/05/disaster-is-inevitable-when-americas-stock-market-bubble-bursts/amp/

[17] https://www.zerohedge.com/news/2018-11-30/how-everything-bubble-will-end

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s