For those who follow Sopra Steria Analytics Sweden and expect the customary end-to-end manual to some model or technique, this post will be a major disappointment as it is entirely devoted to a discussion on the impact of AI on labor and on what could be called “the value of work”. It is my deeply rooted belief that work gives Man meaning and that work has a moral and an existential value. Given this belief, it is reasonable to wonder what effect AI tools could have on our view of the value of work and our evaluation of our being as well as our meaning. We obviously cannot discuss the influence of AI, as it is today and how we imagine it to become in the future, on the value and meaning of work without attempting to define these concepts, however briefly. One could argue for a third value, namely its societal value. What is valuable to society if not every individual’s contribution to its survival? It can be measured in the amount of the burden taken or in economic terms. I argued that both are actually already borne by the moral value.
The moral and existential value of work
It is my deeply rooted belief that every man and woman has a moral obligation to bear, to the best of his or her ability, some of the burdens of humanity (and this is the value of work to society). There is a clear, though unspoken, contract stipulating that we have a duty to assume a part of the work needed for the well-being of Mankind. Doing “one’s part” earns oneself respect and dignity. This is why one feels a sense of exclusion or even humiliation when excluded by the lack of work, of the labor needed in society. The prospect of depending on someone else’s labor is a frightening experience for the vast majority of us, thereof the fears of unemployment and of one’s primary occupation to become obsolete. Apart from the evident necessity of work (whatever the labor might consist of) to participate to the alleviation of Man’s burden, there are other aspects that need to be considered as they participate to elevate every one of us, whether it is by the strength of our limbs or the use of our brains. Work is a teacher and it humanizes us by forcing us to concentrate our effort on very specific tasks. It teaches patience, modesty, courage and perseverance. Patience comes from relentlessly repeating tasks by first failing until the method is understood, perfected and finally mastered. We also learn to be patient by delaying immediate satisfaction of our desires by introducing a certain distance (which we recognize as an effort) between our desire and its concretization. Work teaches us modesty, by recognizing that we are incomplete and can always improve ourselves and what we do. It also takes courage to admit our shortcomings and perseverance to endure repeated failure. But, in the end, we all come out victorious and richer, as well as morally virtuous.
Blaise Pascal (Pensées) eloquently describes what lacking something to do does to Man:
“Nothing is so intolerable to man as being fully at rest, without passion, without business, without entertainment, without care. It is then that he recognizes that he is empty, insufficient, dependent, ineffectual. From the depths of his soul now comes at once boredom, gloom, sorrow, chagrin, resentment and despair.”
The emptiness that accompanies having nothing to do crystalizes the problem of Being, of its purpose and meaning. If boredom and despair is all that is left when we are robbed of our work, isn’t it so that work gives our lives meaning?
Is meaning threatened by AI?
What happens then when this meaning is being threatened? Every revolution in Mankind’s history has raised fears and questions that needed to be addressed and, as we now see the dawning of a new era, that of artificial intelligence, a completely new set of questions see the light.
Aside the ethical questions, which many authors discuss, there is, as I see it, a fundamental difference in th fears of AI and the fears of the machines in previous revolutions: the fear of being completely robbed of what gives life meaning and purpose.
Indeed, all previous industrial revolutions had, as sole purpose, the alleviation of burdens of men and women rapidly breaking their bodies and dying prematurely. These people started working in their early teens, thus deprived of an education and their childhood, unless they were lucky enough to be from privileged families, and died young, some of them while working. Whether it be the invention of new tools, factory lines, medical and surgical aids and so on, none of the advances could replace humans entirely and few imagined they ever could. Today, and thanks to AI, factories can be operated with workers, most transportation systems can be un-manned, AI-doctors can make diagnostics faster (and in some cases more precisely) than its human counterpart. The list can be made as long as the night is black.
Since we seem to have understood that without work, we will eventually perish, the prospect of having nothing to do frightens us. This is true regardless of the task; the question of what constitutes a meaningful work is actually not the issue. The existence of an end result in any work is what gives its meaning; the execution of the work is what gives the individual his/her meaning. We have understood that for a very long time. Already the ancient Greeks knew this and made it into eternal punishments, as in the myth of the Danaïdes or that of Sisyphus, in both cases the punished were to repeat futile tasks.
In this context, the fear of artificial intelligence is quite understandable. For the blue-collar workers, for whom work can be scares in dire times, the prospect of being replaced by a machine that will perform tasks with greater agility is understandably terrifying. For the white-collar, somewhat well educated, being replaced by an intelligent program is just as devastating. For the geneticist having spent years developing a new drug to be replaced by a neural network inventing 20 drugs in a fraction of the time and for a fraction of the price might be an insult, at best. But, in all cases, the fear is there and is very real. The fear is understandable, but it does not necessarily mean it is justified? And, assuming it is justified, what time frame are we talking about?
The past decade’s advances in machine learning and more specifically in artificial neural networks are mind blowing and it is a real exercise in both patience and determination for us working in the field to keep up with all its development. We see very rapid advances and the birth of evermore-complex models and yet, we are only in the infancy of the field. Are we anywhere near the birth of a sentient AI? No. Are we even anywhere near General AI? Again, no, not even close. Machine learning and Neural Networks, today at least, are just another wave of automation and replacing easy, but tedious tasks by a machine. Is this anything new? Of course not. We’ve done this in the past, from the wheel to washing machine, via the printing press, none of which rendered man useless. If nothing else, the printing press allowed the masses to be educated, which meant that the number of literary works exploded. These inventions did not replace humans, they freed humans from tasks that impaired them in their own improvements, that stole time that could be used for more ennobling purposes. Most of the innovations throughout history have been huge forces for the greater good. But, for the sake of argument, let’s assume that we eventually do manage to create an AI capable of doing everything we can do.
Will AI steal all our jobs and leave us unengaged?
Let’s look at this rationally, shall we? Do you know what a daguerreotypist is? What about a hemp dresser? A resurrectionist? A nomenclator? A lector or a lamplighter? Chances are you do not, and even if do know what these people do (because you read a lot), I am willing to bet my life on that you do not know anyone actually having one of these activities, neither as their day-trade nor as a hobby (hopefully). Why? Simply because these jobs no longer exist and as far as I know, we do not have many unemployed lamplighters or resurrectionists. Why is that? Well, we have found ways to only bury “confirmed dead people” and ways to turn on lights without having to physically do so with a stick and a gas lighter. We have, through the eons, developed game changing theories, objects, machines and medical procedures, and not only that, the pace at which we do this is increasing. Some jobs were made obsolete by innovations, some were changed or improved and some were even created. New areas in science have been developed and new ones will be born.
From blue-collar to green-collar
For many young people, jobs such as App Developer, Podcast Producer, blogger, Cloud Architect are all jobs that seem as if they have always existed….they didn’t 30 years ago, some of them not even 5 years ago. All areas of STEMs and SMACs (social, mobile, analytics and cloud) will need huge numbers of people developing solutions. Yes, these jobs will not be available to all and the question of what today’s blue-collar workers will do in the future is of great interest. There are alternatives. Most nations on earth invest billions on, for instance, on renewable sources of energy and the infrastructures will need attendance. The blue-collar workers might just become green-collar workers.
Artificial intelligence also need huge amounts of data to be trained properly. Objects in images in training data, as a easily accessible example, need to be annotated with a sufficiently high level of precision for the AI-solution to be of any use, the data must be made available, cleaned….there are no shortages of things that machines cannot do, at least yet.
We are driven by constantly improving ourselves
Man is not content with what is and is driven by a constant desire to improve. That, in itself is a work, a work on the self. Our history is a witness of that. If it were not so, the printing press would look the same as it did (apart from minor details) when Gutenberg invented it. The typewriter would too and we would be fully satisfied with the functionalities and design of Windows 95, black and white TV sets (if they ever were invented) or the surgical procedures, such as trepanation, and medical cures of the 16th century. We apparently are not. It is simply because we constantly seek to improve things and ourselves. Not only for ourselves, but also for people we have absolutely no connections to. We seek to improve human condition on a global level. This is why we drag people by the hundreds of thousands out of poverty every single day, why the level of education increases globally. In 1820, it was estimated that only 12% of the world’s population could read, while today, only 17% cannot. This is also, why we design artificial intelligence solutions to save lives, the early detection of dangers, save people time to find answers to questions. So far, we haven’t been replaced.
But, let’s assume that we at some point have AI capable to invent things, write books without our assistance and write music. Image that future AI can cook, clean or do any of the activities that we do manually today. Does it really mean that we will not want to do these tasks? You read books that others have written and millions have read before you. Does it diminish your experience of the book to know that? Of course, it doesn’t.
There are endless possibilities for us to develop new trades, new subject to study, new areas to explore. We can even preserve trades and tasks that can be done more effectively by machines or AI. We already do this today. We produce a huge amount of our food in factories, mass-producing exact copies of loaves of bread, sausages, drinks, and yet, the market of the “home-maid”, the “handcrafted” is blooming everywhere. Why would it be so? Simply because individuals value the artisanship involved in bakery, cooking, painting, writing, the uniqueness of a handmade piece of furniture, the love put in the preparation of a meal. We seem to put a lot of our appreciation of personal achievement, whether it be our own or that of others, in the quantity of work and efforts put into reaching a goal, on what one has learned by doing so. We often speak of “putting one’s heart and soul into something” and somehow people recognize this, are touched by it.
In conclusion, the artificial intelligence revolution is not so different from any previous revolution. Some jobs will cease to exist, while some will prevail and gain in value and new ones will be born. Just as it always has been. Mankind is slowly being thrown into a new era, and as always, many questions and new fears see the day of light. The last revolution was the nuclear age and the fear of mutual annihilation, we prevailed. The artificial intelligence revolution awakens the fear of being rendered useless and enslaved. It is my firm belief that we need not fear and that we will adapt, as we have always done. Mankind is resilient. If nothing else, fear will help guide us into harnessing the power of AI, give us safe-guards, help us ask the right question, answer them rationally and make sure that we, as a civilization agree on the dos and don’ts.
We are all capable of great deeds and artificial intelligence is one of them. At the service of Mankind, not the other way around as too dystopian stories go. AI will not depreciate the moral or existential value of work. Will we be out of work? I argue that we will not. Some works will die while new jobs will be invented. We will not cease to think and do simply because AI will be able to do it as well.