You've read some of the headlines about automation that are absolute panic-mongering drivel. It's a serious topic. The mass media knows it and tries to get you fearing for your livelihood.
Some choice cuts, from the front page of Google search results:
"The new white-collar fear: will robots take your job?" telegraph.co.uk
"Afraid of Robots Taking Your Job? You Should Be" thedailybeast.com
"CMV: I'm scared shitless over automation and the disappearance of jobs" reddit.com
"Scary Smart Video Predicts Automation Will Make Human Work Obsolete" mashable.com
"Will robots steal your job? If you're highly educated, you should still be afraid." slate.com
FEAR. STEAL. SCARY. DISAPPEARANCE. AFRAID. SCARED SHITLESS.
But what about the fear of job automation? The fear of "technological unemployment"?
No, that one is not a matter of existential risk. That's a matter of personal job security, DEFINITELY. Which is a problem in itself, but of a drastically lesser magnitude. The effects of large scale technological unemployment on an economy, however, is merely a catalyst for a very simple and (from a humanitarian viewpoint very overdue) change to capitalism.
It's a matter of existential change.
Maybe it feels like a big threat, after the sustained mass media onslaught we continue to endure. The fact that automation is by its nature an agent of change makes it irresistible for panic-mongers. There's even some overlap between the topics of AI and job automation: Artificial Intelligence precursor software already is a portion of the technology humans use to automate work.
Nevertheless, the fact remains that job technological unemployment is neither inherently good nor inherently bad.
The range of tasks that humans can do as work is as broad as the imagination. But all those tasks fit into two categories: Manual labor and Knowledge work.
Making intentional changes to physical matter; Things like printing books, harvesting grapes, assembling electronics, smearing oil on canvas, arranging furniture in a room.
Making new arrangements of information; Things like writing books, devising a wine recipe, writing software, imagining a picture to render in colored oil on a canvas, planning the layout of furniture in a room.
Losing a job you wanted to keep is a problem, no matter how it happens.
Whether all your tasks are more cheaply automated, or you're fired by some hostile, it can be a major problem for the worker.
But a worker losing a job isn't a problem for society as a whole. No matter how certain you are of it. If you're interested understanding the impact to society of the inevitable automation of all work -- and therefore all jobs -- you've come to the right place. We'll analyze every argument espousing automation as a social menace, and you'll also learn about a quantitative solution proposal for improving our primary means of resource allocation in a major contemporary economy, from being "work-based" to a measurably better model. The mathematics prove the financial viability of the solution using current federal data. It's open for discussion.
Assumptions derived from feelings are a poor substitute for a data-based world view. These assumptions don't stand a chance.
Let's get into it!
Assumption 1: "Work is the only acceptable way for humans to gain resources"
Fortunately, the assumption is false.
It's also illogical and dangerous.
The assumption "work is the ONLY way to make income" is an example of taking a DESCRIPTIVE view of the world — in this case, describing that working in exchange for monetary remuneration is for many people the MOST AVAILABLE and sometimes the ONLY way to accrue money — and turning that description into a PRESCRIPTIVE view of a world in which the assumer believes everyone should live that way no matter what. It's an act of observing a situation and then trying to push that situation on everyone else without bothering to ask one crucial question: "is there a better situation?"
Or perhaps the logic is that since the assumer can't think of a better way, then the situation they're in must be the best one possible: "If I'm in this situation, everyone else should be too. I don't even care if there's a better way. We should do it my way because if other people have a better system but I got this worse system, well... THAT'S NOT FAIR ON ME. So we should all get my system so that everyone gets the same deal and that's fair on everyone. And everyone's experience should be consistent with MY experience because I'm already part way through having it and to change my system would be unfair on me."
Despite the obvious logic errors, this assumption stands against any improvement to the system. When a better system becomes available, owners of this assumption fight against it because they value "others not having more than I have" over "people who have the least having more than they do now".
Human work in exchange for resources is our current model in most of the world. That's a description of the situation. I just described it. This system of resource allocation is often called "Capitalism," and there is a lot of evidence suggesting it to be the best resource sharing system yet trialed in a human society for the purpose of resource allocation. Despite that possibility, the system is profoundly flawed, requires strict democratic influence to regulate, and has failed continuously to set a precedent in which all members of a society run on that system have enough resources on which to live. We have more than enough efficacy data from Capitalism in situ to know unequivocally that it is inadequate.
So if our best model is inadequate, what then? We'll modify it. Henceforth we'll refer to the system as "Hostile Capitalism" since it allows for humans to suffer and die for no other reason than having limited access to the resources they need (which is the bad part), while allowing corporations and individuals to profit in proportion to their work (which is good).
When automation makes human labor increasingly obsolete, there will be an increasing proportion of humans who are unable to accrue adequate resources on which to live if Hostile Capitalism is in effect. Logic tells us the system needs to be modified to ensure that inevitable change to civilisation is not calamitous. To be clear, we are talking about the poorest of us having a shot at survival. If you think "The percentage of humans with adequate access to survival-enabling resources" is NOT the primary metric of civilisation, then get back in your toilet. If you think that percentage is the primary qualitative measure of civilisation, or don't know and want more information before forming an opinion, read on for a quick tangent on Income Inequality.
Let's address this important tangent real quick, with another claim and some substantiating contextual points. It's open to discourse if you can improve it, propose a better model, or if you want to dispute the list of substantiating points if your understanding of ethics differs. When doing so, tag it "#IncomeInequality" and hit me on Twitter @autonomike for some concise public discourse. If you've got multiple points to cover on the topic, you can publish an article and tweet me the link. I will definitely read it and almost certainly promote it.
Income inequality of ANY differential is acceptable ONLY IF every human has adequate income and access to the means to satisfy their human needs.
- "Needs" are all requisites for a human life to be sustained.
- All humans have the same general needs, but fulfill them differently.
- Not all needs can be satisfied by money (e.g. community, love, family).
- All humans have different wants, and fulfill them differently.
- "Needs" and "wants" are different concepts.
- Sometimes needs require help to be satisfied: (e.g. intimacy, or a junkie rehabilitating from addiction.)
- Human life prospers when individuals want to help each other get their needs.
- Individuals are generally free to choose to delay, try to avoid, or deny satisfying a need.
- Ensuring all individuals have access to the means to satisfy their needs, and know they do, is society's obligation. Society is not obliged to ensure all individuals actually satisfy their needs, should they choose not to.
At first glance, it appears automation is set to screw up human resource allocation! Automation and capitalism appear to be in conflict.
Fighting automation is not the answer. That battle can never even be waged, let alone won. Stasis can never beat adaptiveness.
Nor should it, since resource allocation among humans need not FULLY depend on human labor. It can in part. Or not at all, as in the ancient-to-modern use case shared by beggars and monarchs.
Resources can be allocated among humans in other ways.
The automation of labor leads to both greater VOLUME of resources (compare yield of a $300,000 of crop harvester to the yield of $300,000 worth of human labor) but also greater VARIETY of resources (compare clothing availability in the period before the washing machine was invented, vs the period after washing machines became household appliances).
You can deduce from those effects that automation has on resources that automation makes the world better for humans to live in.
Fighting capitalism is not the answer either. For there can never be a consensus on the fair distribution of ALL resources. There is no precedent in human history. The "trickle down" of wealth demonstrably doesn't work.
Adapting capitalism is the answer. Remember, all that's needed is a minimum redistribution of resources that enable the survival of everyone. Not a redistribution of ALL resources like in some miserable socialist communist wasteland.
No. A portion of all resources -- like that accumulated in any democracy's tax system -- CAN be allocated fairly, and without altering the fundamental tenets of capitalism. Concepts like Universal Basic Income only require a portion of society's wealth to be implemented. In so doing, all humans gain sufficient resources on which to subsist, thus freeing them from their DEPENDENCE on labour to live.
The cost of this is increased taxation: a trivial price to pay to ensure that all children eat. But what portion is needed? That's a good problem to solve, and none but the most selfish mind would argue against it after seeing the viability of its mathematics. Which are as follows:
2016 Minimum annual living cost for a US adult: $28,474
(We'll call annual living cost "LC")
2015 total population of USA: 321,368,864 individuals.
2015 population of USA over 18 years: 76.9%
Number of US citizens over 18 years: (76.9% of 321,368,864) 247,132,657 adults.
(We'll call the number of of adult persons "AP")
AP 247,132,657 times LC 28,474 = $7,036,855,275,418
(We'll call this cost of Universal Basic Income UBI)
Is this UBI cost affordable for the economy? Can GDP actually support such a cost?
2011-2015 US Gross Domesitic Product size: $17,419,000,000,000
(We call this "GDP")
$17.4 trillion GPD is larger than the $7 trillion UBI expense. So yes, the economy can support the UBI model.
Poverty can be eliminated from the USA at the cost of reallocating 40.39% of the nation's GDP. Today.
US Basic Income as a means to meet basic human needs is mathematically and financially viable right now in 2016
So 40.39% of GDP reallocated to save the impoverished. We'll call this system "Benign Capitalism", because it does not allow humans to suffer and die for their lack of resources (which is good), while allowing corporations and individuals to generally profit in proportion to their work (also good).
In order to transition from "Hostile Capitalism" to "Benign Capitalism" nothing more than a taxation increase to 40% is needed to preserve Americans' quality of life. To achieve this any tax model can be used, as decided by elected officials chosen by the voting public.
The above equation is mathematical proof of the viability of Universal Basic Income as a means of ensuring every American adult's base needs are met.
But what effect would this have on automation?
The effect would be that the owners of companies would continue to profit. Demand for their products would remain generally unchanged, as the very core of capitalism remains unaltered. That 40% of the US's money isn't leaving the economy, it's just moving through a different channel -- one which by its nature guarantees most of it will continue to move, and not wind up sitting static in billionaires' personal surplus. Some producers would see greater profits as a result of there now being more consumers with money to spend in the market. Other producers would see profit diminish, as the higher taxation of their customers affected their ability to spend. The market would do what it always does: balance. And capitalism would continue -- this time with the sustained participation of ALL individuals in the society.
Automation, therefore, would continue. Businesses would be just as compelled to find cheaper automated solutions to labor needed in their production processes than human labor can offer. And humans would continue to work as and when they saw fit. More importantly, workers would find work they enjoyed, and feel less pressured to do work they dislike due to their basic income freeing them from the need to do work they dislike in order to live -- as is the reality for some of the humans in your community as you read these words.
Is Benign Capitalism fair in bipartisan politics?
The social psychologist and author Jonathan Haidt has pointed out that when you speak of fairness to someone with a more "left-leaning" worldview, it means leaving no one behind and creating equal opportunities for everybody. However to someone more conservative the notion of fairness means getting what you deserve and have worked hard for.
Does benign captalism service both those needs? The basic income component satisfies the definition of fairness in which everybody is permitted the means to live. And the capitalism component satisfies the definition of fairness in which reward is directly proportionate to effort: you work hard, you earn more than people who work less hard, same as you do today. Which definition of fairness is favoured more strongly by Benign Capitalism? In fiscal terms we see that the 40% of all wealth being allocated among citizens as basic income means that 60% of all wealth remains available to anyone wishing to work to acquire it.
In left-right political terms, 40% of GDP serves left-wing liberal ideals, and 60% of GDP serves right-wing conservative ideals. That's a 2:3 ratio in favor of the right-wing definition of fairness in which the left-wing definition of fairness is still completely satisfied.
No matter where on the left-right spectrum you put yourself, if Benign Capitalism doesn't sound like a compelling idea for improving the lives of the US populace without political compromise, you can go ahead and viciously subtweet me right now.
Assumption 2. "Some jobs won't get done because some jobs nobody wants to do"
In a resource allocation system like UBI where all humans receive enough money to live on, there will always be those who want more money than the minimum. Perhaps a minority, perhaps a majority. Or perhaps everyone, since humans are innately driven to fulfill their many and various desires. But just as human nature still applies, so too do the fundamental rules of capitalism. There is still manual human labor to be done. There are still jobs to do it that pay money. And there will therefore be humans willing to do the work.
It might be that the work for work broadly considered distasteful would pay more than it did in the Hostile Captialism system; in Benign Capitalism human labor has become a more valuable commodity, now that all humans are free to value their own time in the market without being forced to undervalue it in order to stave off death. Sewer-scraping, therefore, costs more money.
Humans will work for additional resources wherever they value the additional income over a portion of their free time. The labor market persists.
And so too does the onward march of exponential automation.
What jobs do you think will be automated foremost? The ones that fewer people want to do, which therefore cost more to employ humans to do, which therefore yield greater savings to the company buying the automation? Seems likely.
Assumption 3. "The market will collapse because human labor will be prohibitively expensive"
With human labor more expensive overall in Benign Capitalism than in Hostile Captialism, automation will become even more highly sought. Automation itself becomes more valuable.
Consider an hour of the cheapest human labor costing $9, in Hostile Capitalism. Automating that labor to reduce the cost to $1 per hour is a saving of $8 per hour -- at the cost of the human worker who is now in major financial trouble.
Imagine in Benign Capitalism that same hour of human labor now costs $15, because nobody wants to do it too cheaply, and nobody NEEDS to do it that cheaply in order to feed their kids. Automating the same task in the same way to cost $1 per hour is now a saving of $14h -- still at the cost of the human worker, but who faces no major financial trouble.
The comparison shows that while human labor is a now more valuable commodity in Benign Capitalism than Hostile Capitalism, automation is now more valuable as well. Therefore, more automation is required for businesses to maintain profits, and human labor would be most efficiently spent automating labor. Regardless of what happens to the JOB market (in terms of the ratio of human-jobs vs automated-jobs insofar as distinct "jobs" still exist), the LABOR market improves since work is being done for less expense.
Assumption 4: "When labor is done more cheaply by machines there will be less labor available"
When you have possible entrepreneurs in the billions, you have innovation on a grander scale than anything before in civilisation. The result of innovation is invariably greater resource abundance, and greater resource variety.
Assumption 5. "When humans are freed from the need to work, they WILL NOT work"
This assumption is derived from some strong emotions. It implicates the assumer's personal preference: the assumer, as an individual, would choose not work if given the choice. But the assumption is even bigger than that! The assumer has taken their personal preference and projected it onto the personal natures of 7.4 billion individuals whom they haven't met. As blanket statements go, this is one of the largest.
Can the assumption be substantiated?
Can the assumption be disproven then?
Yes, anecdotal evidence at the individual level also applies equally to counter the assumption. Look at any successful entrepreneur. Elon Musk has enough money to live a lavish lifestyle for the rest of his days, yet he busts his ass working for his several companies. Some humans just want to do their work. Some just want something helpful to do. These anecdotes describe behaviours measurable over time, not just feelings over time.
But empirical evidence, of course, is the basis of all good judgement.
Do we have any to validate or invalidate the assumption that humans only work if they need to? Or to show the proportions of humans who will seek out work versus those who would prefer to live off the productivity of others?
Not yet. Universal Basic Income trials are being explored at a national scale in various ways in Europe, notably the work being done in the Netherlands (see citation 8 at the end). Canada and India are investigating and planning their own trials. Even the technology community of California's "Silicon Valley" is getting involved. Efficacy data is near, but not yet here. The knowledge is still being manufactured.
In the mean time we have a hypothesis: basic income will significantly improve human living experience. Such experiments will yield large scale data to empirically show the economic and sociological effects of basic income, whatever they are. The data would then be used to evaluate the adoption and adaptation of the Basic Income model by other wealthy nations. Like the USA, where the economy can easily support it and the only question remaining is: "Should we?"
Until then, assumption 5 remains unanswered.
Assumption 6. "When all labor is done by machines, humans CAN NOT work"
During the process of automating all work, human labor will shift first out of the realm of necessary manual labour (the OPTION of human manual labor will always exist), continuing the existing trend of human knowledge work uptake. In 1920 the ration of manual laborers to knowledge workers was 2:1. By 1955 the ratio was 1:1. And in 1980, the ratio was 1:2. [Citation 9, 10]
That transitional period was a major milestone in the human labor market, and the exponential growth of that ratio has been in effect ever since. The same pattern is being observed with automation. Note in the below graph published by technologyreview.com in 2013, which shows in 2000 the differential spike between US domestic productivity and "job growth". Note also the trend of separation ever since.
- US automation is increasing
- US job growth has stopped increasing and is leveling out
What if these two trends continue? Automation and job growth began to part ways 40 years ago: that's quite the efficacy test. We know we will still need a way to allocate resources, and that working for them isn't viable. We also know that fighting automation is impossible. Benign Capitalism is financially viable and resolves the most pressing issue by decoupling resource allocation from human labor. And it does so without compromising capitalism or the nation's economy. (Yes, I'm clearly trying to sell you this idea. You should buy it, while you still have some disposable income to buy things with.)
Humans will do less work, as technology does more. An increasing portion of the work done by humans will be new types of work that become viable as a result of technology (i.e. the greater resource variety produced by advances in automation). Eventually, the work done by humans will be predominantly innovation: new work, invented to produce resources more efficiently, or to produce new resources that didn't previously exist.
You'll work if you want. And your original ideas will have more value than ever before, thanks to automation.
Assumption 7. "The purpose of human life is to work. When all work is automated, including all innovation, human life will be POINTLESS"
The purpose of human life is to experience joy. Not to pursue joy, and not to create it. Specifically to experience it. Pursuit and creation of joy are definitely part of the means. But the experience of joy is the end goal.
To deny that fact is a hypocrisy. Here's a breakdown of why:
- All human labor has been driven by the fulfillment of a need, the avoidance of a threat, or the pursuit of joy, or a combination of those 3 reason types.
- Every human need ever met, and every threat avoided, was done with the intention to prolong life.
- Every human life ever to prolong itself did so in order to enable subsequent experiences of joy.
From those 3 points we can deduce that every human action is traceable, directly or indirectly, to the pursuit of joy as the end goal.
Consider as many examples as you want. You'll find the logic always holds up.
(By all means cite your clever personal or hypothetical anecdote contrary to these points if you'd like to try to disprove these claims! If you believe the prime directive of human life is NOT to experience joy, it would be valuable to us both to understand the basis of your belief. It would be interesting to know if any of those claims were wrong.)
Regardless of our beliefs, the fact remains that once all work is automated there will be nothing left to humanity BUT the pursuit of joy.
So we'd better make sure our descendants can handle it.
Thanks for learning!
And this punchy conversational resource by the genius Eliezer Yudkowsky, who reminds us that labour isn't a finite lump; the scope of possible labour literally matches the scope of the aggregated human imagination.
Or broaden your world view by exploring the citations and data that contributed to this article, upon which this proof of concept for Benign Captialism was based. Inspect the merchandise, so to speak. I'm sure you'll find it all in order.
CITATION 1: Cost of living calculated for the United states, specified by CareerTrends.com http://cost-of-living.careertrends.com/l/615/The-United-States
CITATION 2: 2015 Population of USA, specified by CIA World Factbook: https://www.cia.gov/library/publications/the-world-factbook/geos/us.html
CITATION 3: 2014 USA Populace under 18, specified by US census data: https://www.census.gov/quickfacts/
CITATION 4: US GDP, specified by WorldBank.org http://data.worldbank.org/indicator/NY.GDP.MKTP.CD
CITATION 5: Why GDP is the most valid metric for measuring the size of a nation's economy, specified by Investopedia.com: http://www.investopedia.com/ask/answers/199.asp
CITATION 6: http://psychclassics.yorku.ca/Maslow/motivation.htm
CITATION 7: https://en.wikipedia.org/wiki/Maslow's_hierarchy_of_needs
CITATION 8: http://www.theatlantic.com/business/archive/2016/06/netherlands-utrecht-universal-basic-income-experiment/487883/
CITATION 9: http://www.nickols.us/shift_to_KW.htm
CITATION 10: http://forschungsnetzwerk.at/downloadpub/knowledge_workers_the_biggest_challenge.pdf
A data driven look at the history of automation from The Guardian https://www.theguardian.com/business/2015/aug/17/technology-created-more-jobs-than-destroyed-140-years-data-census