Job revolution (I)

Introduction

Technology, science and new jobs. The future of the labor market is now in the range of possible discussion topics between the bar talkers. That topic is in the news and in the main massive media, but what is the true and what is not about they said to us? Their affirmations are usually contradictory. (I do not want to enter in the alternative topic of discussion as it is the relation between the communication journalism and the technology and science dissemination, we will have time in the future of this blog adventure).

The purpose of the present post is to share with you my based opinion about all that topic, using scientific publication, media news about important figures and other material.

The first we have to learn in order to understand that topic is to be able to notice the difference between a ‘job’ (in Spanish ‘empleo’) and a ‘work’ (in Spanish ‘trabajo’). In the common use of the language, both words are used interchangeably (in different languages happen also, e.g. spanish). Here we are going to select some specific meaning of each word:

• The word ‘work’ for a task, an activity.
• The word ‘job’ for an economic role for which a person is paid. Some works can be jobs but not all of them are jobs. On the other hand, all jobs are works within that definition.

The possibility to do a work is only constraint by the physics laws or the human laws. The possibility to do a job is constraint by the relation between the technology, the society and the economic system.

Once we have set the terms and the words with no fuzzy definitions in order to avoid sterile unproductive debates or semantic debates, we are able to start debating and understanding the job dynamics topic and some common errors and misunderstandings about it.

Science, technology and work

The first we are going to explore is what is a work and which relation has with technology and science. In order to don’t lose track of the main topic probably are going to be too dummy, but the importance is to get the idea. We have to say that the range of possible works is huge and the relation of each of them with science and technology are different. The ones based on enhance science and technology as innovation or research have totally different relation with technology than the ones based on only taking profit of this new technology but not developing them. (This split classification is not always possible, because sometimes we have works related with both of these classes, but for this dummy exploring it is OK).

These both types of works we just defined have different natures and in consequence, different interactions with science and technology. Taking profit of technology it is an incremental process. Technology and science do not forbid you to do the thinks without considering these new technologies. We still have the opportunity to plow or mow the grass without considering any machine. For these kind of works, technology and science only creates new opportunities, new possibilities to do work. It is always a new producer of new works, because it is not able to destroy possibilities of working.

On the other hand, innovation and research are highly dependent on the state of science and technology in that moment. It is difficult to defend that, in order to discover some specific theorem with the same mathematical tools and the same path it was discovered years before and publicly available for everyone, it should be considered as a work (in case of considering also that possibility as a work, the nature of the relations between these works and the science and technology will be the same as the type of the works that takes profit of them described previously). In the case of not considering that a work means that it is possible to ‘destroy’ possibilities of work. But each new innovation opens a range of new possibilities for future development of new science and technology. Probably, in far future (see Superintelligence and technological singularity) years, that fact do not happen but since now we know that science and technology is a net producer of new work. But not only a net producer, indeed it is a very good net producer, the balance is very positive for creation of new works. So, we even can say that science and technology are a strong source of creation work possibilities.

Science, technology and job

The relation between science and technology with jobs is a little bit different in comparison with works. Now we have to consider society and economic system to understand which works are jobs.

For the sake of simplicity, initially, we are going to consider a strong different and opposites examples of economic systems as they are the economic liberalism (as capitalism) and statism (as communism). We will try to understand why jobs are highly dependent of the economic system. And in order to do it easy, I am not going to enter in the small and complex details of these economic systems (for the understanding they will be enough as examples). In order to reduce semantic discrepancies between us we are going to give and easy (and dummy) definition of both systems.

• Statism as a centralized system called state (in which it is aggregated the conscious opinion of the people somehow) in which the works that are jobs are defined by the state following some defined criteria.
• Liberalism (economic) as a decentralized system called free-market (in which it is aggregated unconsciously the opinion of the people following the market rules) in which the works that are jobs are defined by the interaction between the society interests and the technology. The real world is neither one or the other but a combination of both of them.

The definition of job by the statism is easier for our understanding. In some way, the State is able to define consciously which works have to be jobs (because of the common interest for the whole society for example) or not. The decision follow some global criteria or some individual will or believe (elites wills). So, it could be arbitrary depending on how it is decided the criteria and how it is politically defined the State and it is less flexible (or stable depending on the problem and the point of view) to the small changes in the interests of the people and the changes of the science and technology.

In other hand, we have economic liberalism in which people have some assets (money, time, materials…) and they decide how to use these assets considering some personal criteria (following some material, moral, social incentives, a complex combination of them or by rolling a dice, whatever) and the possibilities the system allow them. Some works even they could have product demand could be or not be jobs depending on how are their net balance production in comparison with other jobs, or if it has an enough job demand to reach some critical job employees in this sector or a critical demand in the local region, between another conditions out of the actual direct will of the people. A drawback (or a good point for others) is that even it claims individual freedom sometimes the market imposes or enforces social incentives by ‘majority get all’ events. On the other hand, it is more adaptable to social changes and science and technology changes (but less stable) and sometimes it lives too much in the present, forgetting the past or living without any strict planning about future.

History context

In order to have a global picture of jobs evolution we are going to travel through history. At the beginnings of the human live, we lived in a pre-market society. There was no money and probably no direct exchange of goods, so no jobs as we defined at the beginning. The people lived in tribes and they shared most of the goods and the tasks that they had to do in order to survive (hunting, cleaning, cooking, defense,…). In exchange they were indirectly provided by protection, security and some sort of stability. No strong interactions (in time criteria) with other humans out their tribes. With the beginning of the civilizations, the people tended to interact with more human beings and the property arose. Also it appeared direct exchange of assets, or jobs we defined. But it was not until the appearance of money (and other technologies) that the trade of goods was improved, allowing people to specialize themselves. In consequence, major part of the people forgot most of the surviving works (these works directly related with the basics of getting food and keeping home conditions in order to live in a stable environment). Now there were able to buy houses, food, basic utensils to eat or dress themselves. They still had to do basic surviving tasks as buying, cooking or cleaning but these were done by women or child. In that moment labor market and jobs as we know now appeared.

A massive change in production of utilities, destruction of survival jobs, substituted by jobs in companies and factories and another works as buying. Thanks to the new machines and technology a lot of jobs related to agriculture were substituted by more productive ones but the ratio of agriculture jobs decreased. The arising of money economy lead to other jobs in manufacture. All these events, which led to a change of paradigm in labor, were called by the historians as the Industrial revolution. A strong transition between works and low-scalable jobs mostly related with agriculture and basic tools made by craftsmen to industrial production and money-led economic system. In that times workers had 12-14 hours working day. The machines which produced this transition only were able to be used by humans to enhance their performance and productivity for only low-level, routine and repetitive tasks. Since then the labor market was able to reduce the hours of working day and incorporate most of the women.

The automation leads the labor market into service sector jobs instead of the manufactures. As it was moved jobs from 1st sector to the 2nd sector in the Industrial Revolution (helping to escape from the Malthusian Trap). This next revolution (for some ones called the information revolution), not only is moving jobs from 2nd sector to 3rd sector but also it is turning fuzzy the frontier between jobs and works another time (cooperative economy or other alternatives are exploring the border). And some questions arises. Which is the important factor to keep jobs market alive and able to give jobs to everyone? Liberal market as someones defend? Intervention in the economy as others said? Or more complex factors as a good balance between adapted offer of services with equality distribution of rewards? These questions are very difficult to solve now with the information available. And I save most of them for future blog posts. So, solve that questions is not our goal here, today, and some researchers are dealing with it (and leading opinion as the AAAS researchers). In that section we are exploring history and the main problem is the access to meaningful information to bring us over to some possible conclusions (past data and present experiments).

That is the reason why it would be interesting to study the evolution of the jobs along history but knowing all the jobs (better than in number in hours of labor) of these times it is a challenge. Economists at the consultancy Deloitte tried to do that collecting job data from UK (in number of jobs unfortunately). The study is good to try to understand past dynamics. They told to us that there was net creation of jobs since the Industrial Revolution. Most of the jobs created are the more intellectual tasks. Most of the jobs destroyed are the ones related with the routine tasks. With that data is difficult to reach more conclusions.

Forecasting job market changes

The new step is in the future. How is going to develop the job market? Related with that question I have to make some complains. I have a complain to do about all the job forecasting studies I read. All of them have the same bias aptitude as regards the paper of society will and the States in the future of the job market. As the only thing we should study, in order to understand the future of job market, was the new technologies but not the society or our economic systems, defining them as reactive elements in the technology development as an assumption.

This reactive assumption probably has as an explanation of keeping simple the models or because of their believe that the past social and political situations did not affect in the course of science and technology history, in which sentence I can not agree.

As an example of bad studies I have to mention the one from Deloitte cited before. It is a good study as a description of the past dynamics of job market but it is not for prediction for some technical disagreements and complains I have to expose here:

• History do not only follow cyclic neither homogeneous dynamics. And even less for complex features and questions as job dynamics is, which depends on technology, society and economy interactions. The in-lines assumptions of principle of mathematical induction for history do not talk to us very well about that study. Facts in the past could or not be directly translated into the future. Situations in economy and society changed. That is not a prove of wrongness of conclusions but of the methodology. We need to study better what are the static patterns through the history, not assume neither impose them.
• I agree that most of the studies are also biased because of the balance known-unknown, as they said. But it is a strong assumption that all new works created will be jobs (we agree that there will be new works, as it is exposed in the introduction of this article).
• Some jobs created are because of the State decisions or with strong involvement of the State (as nursing in the study). We can not say if that situation is a reaction or not, but it is not settle as the study assumes by default.
• As the study says, there are some of the routine factory jobs that were place in other countries because of globalization and cheaper workforce. They do not say that others, like actors, could have increased in UK because of external market also. So, without a whole picture it is difficult to decouple territorial dynamics. And so, to get founded quantitative conclusions.
• Generalizing partial results could be dangerous, and all the more so if we bear in mind that this partial results are from a difficult representative of the whole population of cases.

But most of the studies make the same methodological mistakes. The quality of the data is not the desired, the constraints are high and they usually try to make predictions with mistaken tools. Making strong assertions with a low quality of data, methodologies and results could be misleading for press and people. We have to be aware of that situation.

So, if we want a good predictions, it is not enough with only past data neither with dummy computations about the marginal productivity of some automation tasks. We have to take care about social changes or economic system changes, the probable reactive ones and the possible active ones.

So, we have to resign ourselves to study that problems? Of course not. Due to the complexity of the problem, the low quality of data and methods, and in consequence, with the difficulty to tackle that problem quantitatively, we should try to be less ambitious and more realistic with our possibilities.

Even they are not conclusive, it could give to us some insights. A task which is cheaper do it by a machine rather than by a human, without considering future taxes, it could be in danger the next years. Political protection of these jobs could delay their destruction, but without any special situation or social contribution of these tasks, it is difficult to defend their survival. Some shifts in peoples interests could lead to other services jobs. Some existing but mediatically covered problems (as it could be suicides in front of tobacco or car accidents) could be a source of new job positions.

But what methodology we have to follow? Which are our goals? Without any evidence or clear forecasting power we can only talk about probable possibilities and concerns. In order to study the problem we are going to use mental simulations by imaging possible situations.

In that post we are only going to study that problems in the high-level. We will have other opportunities to seek deeper into them. The main idea we have to have into mind is that machines will be able to substitute all the possible tasks we are able to do today. They will be able to do creativity work. There are studies that creates machines which are able to create paintings or compose songs. The typical concern in computer science about the problematic P=NP lost their importance, machines now try to tackle NP problems by probabilistic computation (as humans do), they are improving a lot.

We have some examples of possible changes that the incoming technology will probably produce:

• Small decisions taken by machines. Small decisions like what we want to take for dinner will be taken our conditioned by a machine opinion.
• Cooperative economy, sharing wastes and taking more profit of the assets we have. That will blurry the lines between doing a commercial activity or only sharing. Governments will have problems to regulate that and apply correct taxations.
• Reduction of the proportion of job hours in front of work hours, and routine tasks could produce or a reduction of the working day or a bad distribution of jobs.
• The solutions to bad distributions of jobs could lead governments to force reduction of working days, implementation of possibilities of a highly diversified labor hours contracts (contracts with a lot of different and unconnected jobs with short learning curve, as customer services or others). That could allow people to work in 2, 3 or ore different jobs in a week and only few hours (as working in a customer service, teaching children, and in weekends playing guitar in a bar). The way to do it could lead us to inequality. Probably most of that economy should be run by social incentives like new experiences than by a material incentives.
• Shifting routine jobs to artistic jobs could create inequality. Artistic world is one of the sectors with worse income distribution. There are few artist that take all the market and other with few visibility. There are possibilities of changing because of the shift in peoples interest and the shift in the economy to the services sector which will put more money, not only in the sector by itself, but also with better redistribution. That is the idea of the pro liberal market, but it is not needed to be in that way. And we should wait to be if it is enough to redistribute well the money in that sector.
• Machines will do most of the jobs of our economy. But there have no rights to earn money, so their owners hold the capacity to earn that money. We should remember that the mobile phone is a machine and it is giving to us a lot of possibilities with a cheap price. In other hand most of the manufactures are fill with a lot of robots also, but not as cheap as the mobiles. For economies ruled by cheap machines, it is more probable to have a good distribution of wealth. It is (materially) easy to enter in the market and the earns will be shared easier. But with expensive machines ruled sectors the monopolistic risks arises. We should wait to see if there will be closed ruled markets or not to say if the market economy will increase in wealth distribution or not.
• Social computing could change the way we understand the organization of our democracy, with the creation of new problems and solutions. Social computing could make easier for society and organizations to organize public sectors without a State. Fluid democracy could come with flexible collective movements and commitments. That new fluid democracy could lead us to shared bus transport voted by people and organized by people without an intervention of some public government with the good and bad points which that situation can cause.

So, our main warnings in the future will be classified here as:

• Ethical problems (changes in the moral incentive balances). New technology will change the way we live and interact with other people or our tasks, as we exposed in the last itemizes. We will delegate most of the routine day-to-day decisions, but also we will make our life more unpredictable. We are what we do. If we are able to combine possibilities our life will be able to differentiate more from the others and exploit our personal strong points. About that topic my mind always recall the American The Office TV-show in which the characters watch their live flow, conditioned by their office situation, and confining their relations in the domain that the office offers to them. Freedom is in discussion, more unpredictability will make governments, corporations and machines less able to control the incentives of our lives or make you (without notice it) to take some specific options. We should remember that manipulation is that, manipulation is not saying or forcing the people to do something but controlling their incentives in order to make them take that decisions thinking that the decision is the one they wanted.

• Economical problems (changes in the material incentive balances). We have a competitive economy in which the people who is able to produce something which more people wanted is able to earn more money. In the past, technology helped people to do their tasks better, reducing the cost of the survival needs and increasing living standards. In the future, the machines will be able to compete directly with people in doing some tasks completely by themselves. Even without earning money machines will run most of our economy (indeed now big part of the stock market is run by machines). Besides the ethical problems that it will carry, we have different possible situations:

  • Machines taking material decisions. Deciding who is going to earn money or not an participating actively in the money transactions. That situation can lead us to a supra-economy mostly run by machines (and for humans and machines) and a sub-economy, mostly run by humans (and for humans).
  • An economy lead by humans and for humans. The machines actions are regulated for the sake of keeping some control over our economy. The humans have most of the survival tasks covered easily (because they are provided by the State or/and cause the technology is cheap) and the rest of economy transactions are for extra consumptions, experiences and caprices.

In both cases there will be two main different human possible jobs types: - Quick learning curve, or that require living skills knowledge: almost centered in giving services for humans (providing experiences, art, human interaction, psychology, entertainment…). People with more diversified working times and probably more unequal distribution of wealth. - Slow learning curve. People needs to be specialized. The diversification of working hours will be more difficult. Most of the work centered in science, technology innovation and translating that work into machines improvement. The wealth distribution probably will tend to be better for the decreasing of the costs in technology and the high dependence on the work of others.

• Social problems (changes in the social incentive balances). Changes in the way we interact will cause a changes in the social incentives. Now, the arising of the social networks and the high-interaction between people caused some problems. Social incentives became strong, and making some bad decision could make you famous and massively attacked. You also have in power and incredible way of communicating with people or even influencing them. Also you are seen by others and your Internet profile can be decisive in your future professional career.

Conclusion

So, technology, machines or robots are not responsible of stealing our jobs and yes, the Luddites movement have no sense. If there is a net job destruction because of new technology the only conclusion available is that we have an economic system which is NOT able to fit with this new technology developments.

And of course, we have to be cautious with these forecasts. Economic systems can change and if some forecast is highly dependent in some not considered variables for the study, these forecasts will fail considerably. Which it really seems to be more predictable is the balance between the incentives in the decision making on most of the people. Since technology was developing and influencing in the people, society and labor market, most of the works which was able to be done by a machine was depreciated. People stop worrying also about only their functionality or other material issues and started worrying about more aesthetic, social or even moral issues. Some of them were able to be transformed, change the market model into focusing the business model in other non-material incentives (aesthetic, ecological friendly, socially attractive, manually built,…) but they lost their market scalability (no massive production of them).

That also explain some lost of interest for the material properties and the increase of the market of experiences (traveling, experiencing extreme feelings, extreme sports, taking a coffee in a very select environment and live music, …) So, we have to be worried about new automation of tasks or not? We have to go to more centralized system or a more decentralized one? Which consequences will carry out this new labor revolution?

It is difficult to get into all that subtopics in this article. We will recall some of them in future of this blog adventure. I save these new topics for a new future entries in this blog. See you next time :)

And remember, we are living great times for science, technology and culture. Let’s enjoy it.

Bibliography

For more information:

• Gregory Clark (2006), A Farewell to Alms: A Brief Economic History of the World. Princeton University Press
• Boldizzoni, Francesco (2011). The Poverty of Clio: Resurrecting Economic History. Princeton University Press
• Douglass C. North (1965). The State of Economic History, American Economic Review
• Ford, Martin, (2015). Rise of the Robots: Technology and the Threat of a Jobless Future. Basic Books
• Robots are coming for your job … and that’s not all – Tech Weekly podcast, TheGuardian