Robot vs. Human
Have you thought of competing with a robot for promoting within your company? Although people would be better in lobbying, considering costs and performance it is possible that you’ll be surpassed by a robot to become the manager of your department.
Just consider going into interviews with robots, would you ask them where they see themselves in 5 years? That’d be even worse if they are the interviewees asking the same to you and having the processing power to estimate if you are reasonable or not.
How about a ‘mobbing’ robot manager? It is programmed for mobbing in the workplace so it would be rather difficult to cope with it. Who would you blame, the engineer or the robot?
These may sound a bit far-fetched when we consider the current status of automation on the work floor. Nowadays, and for some time, the issue has been the fear of people for losing their jobs because of increased automation levels. International Federation of Robotics estimates that by 2018, around 1.3 million industrial robots will be in manufacturing lines around the world. Does this really mean that minimum of 1.3 million jobs will be lost?
Actually, it is not trivial to conclude that more automated manufacturing lines can result in less job openings on the work floor, which is referred to as technological unemployment. It is complicated because of the fact that job opportunities in different levels will be compensating the job losses. On the other hand, decreased costs due to higher productivity may result in more demand of certain services/products, therefore new jobs and even professions may be created thanks to automation. Not to add more speculations to this issue, in this article, we will be looking at the roles that robots will take over from people and which skill sets will be necessary to win this competition.
For a good understanding of the future of the technology, it is wise to look at the very beginning. So, the first step is to look at the definition of robot and what roles have been covered by robots so far. Robot, as a term, first appeared in the play by Karel Capek called ‘R.U.R.’ around 100 years ago and it is derived from the Czech word ‘robotnik’ meaning ‘forced worker’. Although the inventors of robot arms have not used the first example as a ‘forced worker’ but as a bar tender, that inspired industrial robots and indicated the possibility to use machines for repetitive tasks.
Later, robot name is given to any electro-mechanical machine driven by a software program, which is capable of performing tasks autonomously or semi-autonomously. The very first industrial robot appeared on the work floor, back then called Unimate, in 1961 at General Motor’s factory to lift hot die-cast metal parts. They were introduced in automotive industry for welding, painting and applying adhesives. (1)
At the first place, in order to increase productivity of a factory and thus profitability of the businesses physical manpower on the field is replaced by robots and they were taking over tedious, repetitive, highly structured tasks from people.
Since the introduction of automation, people have been afraid that technological developments would cause increased unemployment especially in low level jobs. As an example, during Paris-Roubaix stage of Tour de France in 1976, workers of Parisienne newspaper held up the race for an hour to protest against the introduction of automation in their printing facilities. This fear proved to have grounds according to some studies (such as Jeremy Rifkin’s long discussed End of Work book) and some other views claim that it has lead to other types of job opportunities for which other skill sets are required.
Apparently, the problem was not that people would love to do tedious jobs but it was the fear of losing jobs and even worse security for their future. So, we have been struggling with a fundamental concern which can be related to Maslow’s hierarchy of human needs.
According to Maslow’s theory, once the basic needs are satisfied people would struggle to sustain their stability for the future. Supposedly, the automation/robotics technology would have been challenged by safety and security concerns at the first stage and resulting in resistance against automation at the work floor. Indeed this was the case, which resulted in several solutions for improving the security feelings of people and securing the position of automation technology in the market. Politicians have proposed several ways such as basic income (2), shorter working hours (3) which are still under discussion and even trials have been done in several countries.
Recently, we have already started seeing unmanned operations or facilities running with minimum human interference. Typical applications of industrial robots include assembling, dispensing, handling, processing (for instance, cutting) and welding, all of which are prevalent in manufacturing industries; as well as harvesting (in agriculture) and inspecting of equipment and structures (common in power plants). (Robots at Work, 2015) Declining costs of robots will make this transition even faster and more widespread in different industries at different scales.
A famous example of this is Amazon’s fully automated warehouse, employing (!) 30.000 robots in their so called fulfilment centres. This level of automation is required because of the severe conditions that Amazon warehouse employees have been exposed to. It would be interesting to see robots revolting against the ‘subhuman (or sub-robot) conditions’ in the warehouse.
Another example is Rotterdam’s Europoort, not only one of the busiest but also one of the most innovative terminal around the world. Container loading and stacking in the port is almost fully automated. People are working in the monitoring and control rooms where one person can handle minimum 10 operation lines, which was 1 for 1.
In these facilities, people are working in engineering, operational and managerial roles. So, there are new type of jobs generated for us, more in the managerial direction. On the other hand, technology developers are looking for ways to find possible cooperation between people and robots such that they can still work together; cooperative robots, namely cobots, which are able to sense their environment and also colleagues are attracting attention. Basically, the connection has been established between humans and robots for co-existence on the work floor. Currently, experts are trying to define standards for this cooperation and this is going to be the agenda of engineers for some time, coming decade possibly.
These are the indications that we are in the transition period of the technology development in automation, being connected and cooperating with people; trying not to compete. On the other hand, having accepted the existence and necessity of this technology, people are looking for ways to develop complimentary skills such that robots will be supporting people. So, we are still climbing in the pyramid of Maslow: dealing with self-esteem and unique positioning inside the industry. People will be developing skills to secure their position and in respond to that technology will be developed for their existence.
Once positioning is set, then we will observe the next upcoming phase of technology development in reference to human needs pyramid. Accordingly, one would expect to see moralities and values will come on the stage and actually, yes, that’s the case.
As indicated above, we can analyze the future of this tension in two phases: the requirement of unique skills in near future and moral/value driven technology development in relatively further future.
At the first place, we should be checking which skills will be necessary and what kind of jobs will be created. For that, it would be wise to look at the current trends in the job market.
It is trivial to see that people are moving from physically active work floor to mentally active office floor; the demand for crafting skills is diminishing while demanded qualifications are getting more challenging to meet. People are expected to observe and think, becoming more and more mentally challenged. Today, in October 2016, on a job board searching for a technician vacancy in the Netherlands delivers approximately 10.000 results. Whereas the number of engineering vacancies is around 16.500 (and the biggest share devoted to SW engineers), administrative vacancies more than 19.000 and manager vacancies almost 33.000. (These are fetched from indeed.nl and the keywords: monteur, engineer, administratief, manager)
Currently, management and administration roles constitute the biggest share of the jobs on the market and our quick search is inline with the research of European Employment Services, with a slight surprise that currently artisans are the most requested work force in the Netherlands. (4) However, this is likely to change; computerisation in administration, logistics, transportation, sales and service is expected to cause substitution of people with computers in the coming years. By the introduction of computers, resources for low skill and low wage occupations are expected to be allocated to jobs that require creative skills in the future. (5) (6)
Looking at the trends in human resource management (HRM), companies outsource their talent sourcing efforts and buying services from HR professionals to find right talents for their teams. So, understanding the skills required for a job is getting more complex which is an indication of more specific requirements of jobs. Many studies regard talent shortage and resourcing highly skilled people as one of the main challenges in the future. For example, PwC’s yearly survey with CEOs reveals increasing concern for the availability of skills every year. (7)
Considering the resourcing challenge and the need for reallocation of resources from low to high skilled jobs (highly probable for computerisation to low probability in the image), HR professionals are expected to deliver solutions for training the current and retraining the future work force. At this point, in order to determine the crucial skills for the future, one should look at the prioritized roles in the future as well.
What kind of jobs & skills will be the top priority in the future?
This issue was addressed in the last World Economic Forum under the topic of Future of Jobs and the concluding report reveals very interesting (but not surprising) future perspective. (8) According to the report, not much different than what we presented here, architecture and engineering jobs will be on top of the list while office related administrative functions are expected to decline in the very near future. The market will ask for employees with cognitive skills such as creativity, logical reasoning, problem solving and social skills such as coordination, communication with others, emotional intelligence, adaptability. These are complementary set of skills on top of basic skills to perform our jobs. Therefore, this implies higher and multifaceted education.
Basically, we will be required to have higher education levels in the coming time. This has been already the case such that the average level of education has risen up globally (9) and in the coming years it is expected to rise further. So, in order to be able to get a unique position inside industry and even in our modern society, people are pushed for higher education. We shouldn’t be missing an important point here: We are not talking about traditional education in one specific area of expertise but a multifaceted education in interaction with other disciplines and even with art, philosophy.
Many ’10 jobs that didn’t exist 10 years ago’ lists suggest that analysts are highly demanded in the market nowadays and that’s actually a proof of the predictions above. For many businesses, people have enough data/information available today but it is a matter of compiling the ‘right’ piece of information together and create solutions with multidisciplinary teams.
As we consider engineering skills necessary for future jobs, being excelled in one specific area will not be enough. Technical experts will be (and actually are) required to operate in multidisciplinary teams, which urges communication and understanding of more complicated problems in higher levels. As an example, a mechanical engineer assigned to design a mechanical component of a robot, cannot focus only on the mechanical design of the component but interfacing with other units (such as electronics cards, control units, actuation modules) is critical. It has been always critical but these systems are integrated much more than before and more importantly it is possible to model/analyze these interfaces earlier in the design stage. Therefore, any design decision of an engineer cannot be taken by one but it must be taken in consultation with others; where the social interaction comes on the stage. So, engineers of our age are also socially challenged and this challenge is going to be complexer. Is it better to have a software for taking decisions in these circumstances (avoiding mistakes because of communication inefficiencies) or is it that interaction (and inefficiencies) which makes people creative to think out of the box?
What about the skills of robots in the coming future? Engineers and developers have been already busy with making the robots/automation technology unique by adding skills of learning, right decision making on top of sensing the environment and data communication. For example, ’deep learning’ has been a hot topic for achieving higher levels in artificial intelligence, which is aiming to bring artificial intelligence closer to the level of human intelligence. So, for robots (or any other automation tools) flexibility, independency and reliability will be the key skills which are not far from our technological reach. However, there are great deal of work for developers (and if AI is self developing, for AI) to achieve creativity at human level.
What about the power of decision making?
If we are talking about creativity or out-of-the-box thinking, we are already one step further in Maslow’s pyramid. The question at the next level of development is if decision makers’ roles will also be occupied by a software or a robot. It is not anymore only efficiency in physical tasks but the decision making process which makes a business profitable. Especially in today’s fast paced world, it is even more critical to take right decisions on time. Nowadays, computers and software are already assisting people in decision making processes by both data analysis and future predictions. Processing power is essential for estimating the consequences of decisions and considering the complexity of today’s challenges. Therefore, we may see software tools taking over these key roles as the system grows bigger and decision making becomes more complicated due to huge amount of data and also driven by data.(10) However, the challenge is much more fundamental than only feasibility, it comes down to moralities and ethics.
Although the trend seems to result in handing over our decision making power to robots and software, we cannot yet rely on tools, even each other, especially when it comes to decision making in critical points of development, operations or management (and indeed even in our daily lives).(11) This is similar to the long discussed problems of autonomous vehicles; people would like to keep the steering wheel still in their own hands. If it is only about processing power for analysing available data, those days are not far to see computers in CEO chairs. The processing power will surpass the capacity of brain of mankind at affordable prices, thus we will be forced to let ‘them’ do it. About 20 years ago scientists predicted that to happen until 2020.(12) Although it is not yet scientifically proven if that’ll be possible, some influential leaders in technology and science seem to be worried about the upcoming challenges of AI takeover.(13)
Nonetheless, right decisions are not only result of reasoning but also intuition. The important role of intuition in decision making process of experts is widely expected and there have been many stuides aiming to understand the intiution process and compare it to rationalised decision making process. Whatever the results are, it is and it will remain critical to engage intuitive power on top of processing power. Of course, we will see engineers pushing the technology to introduce intuition to software and robots. While we are having difficulty to explain how the intuitive process is exactly working, it is definitely a big challenge to have ‘intuitive robots’. More interestingly, if machine learning can lead to intuition process, in relation to pattern recognition in data, we can see that artificial intelligence can take intuitive decisions as well.
Autonomous systems are on top of the list of technology development and it started raising new discussions around ethics, namely machine ethics or more specifically ethical robots. This has been a hot topic in many areas ranging from medical and transportation to defence techonologies. For example, lethal autonomous weapon systems (LAWS) are developed to exclude human out of the process of selecting and engaging targets; some argue that once the ethical rules are set machines are much more reliable than human when it comes to following codes. However, then the question will rise: Who will set the rules? In case of manufacturing automation, ethical discussions seem ‘less tricky’ but it may not be the case. Imagine the decision of manufacturing volume to be decided by an automated system while it may result in laying off employees or dismissing long-time suppliers, would it be possible to let an automated system decide on its own? The point is not to say people are morally better (indeed not), it is just that people are more flexible, empathic and negotiable compared to machines.
Coming back to Maslow’s pyramid, the need will be justifying our existence through moral values and creativity. At one point, the dilemma will be to decide if we can rely on machines not only for the sake of right decisions but also for creative and ethically right decisions. Even if all the above mentioned issues are resolved, full autonomy entails accountability and at critical points this will be a difficult discussion to find out who is responsible for the consequences.
Concluding remarks and more questions
Having considered all these points, we ended in more questions rather than answers.
We tried to summarize where this technology shift started, presented the current status and the near/far future where it will lead to. this whole automation idea is built on the fact that we would like to reduce the costs by increasing productivity and save time of people for more valuable activities. We expected that increased productivity would mean higher economic value and more jobs. However, following figure implies the vice versa. Once the possibilities are discovered, this genuine reason seems to be forgotten and we are starting to compete with the tools that we created ourselves. Being the creator of this technology, we are shooting our own foot for the sake of profit. So, the questions will be: Where will we stop innovating? Or in a different way, is our society able to keep up with the rapid change in technology and respond to it by creating more jobs?
We explained the technology development and reaction of society to this inline with Maslow’s pyramid. Using this approach, we estimate that unique skills will be necessary for both robots and people to co-exist in the industry. We spotted complex problem solving and social skills to be key skills for the future. However, these discussions may be obsolete if programmers will be able to equip a robot or a human (or a cobot or may be romployee) with necessary skills for a certain job, for example uploading tool capabilities, language skills or necessary know-how. Should we worry about the next moves in the job market?
Following these, Maslow’s model suggests that we will end up in ethical and moralities discussions which we are already engaged with and the question will come down to: Can we replace human creativity and morality by robots’? will technology allow artificial intelligence to take the role of human beings only by reasoning? Or will it be possible to teach intuition to machines?
Our answer to all these questions is simple. Yes, robots/software will replace human power for certain type of jobs and they will develop many skills themselves thanks to Artificial Intelligence. They will replace people in industry in critical positions as well. However, the adaptive power of human being is very difficult to compete. This power is also based on skills of human beings: intuition, curiosity and creativity. These skills make people inconsistent and inimitable by algorithms even if any system may have more processing power than human brain. Precisely, the inconsistency of human decisions make it extremely difficult to model and compete with it.
2. For further reading about basic income (from a pro-basic income view): http://basicincome.org/basic-income/
3. A Strategic Opening for a Basic Income Guarantee in the Global Crisis Being Created by AI, Robots, Desktop Manufacturing and BioMedicine: http://jetpress.org/v24/hughes2.htm
4. http://www.loonwijzer.nl/home/carriere/waar-vind-ik-een-baan/meest-gevraagde-banen (in Dutch)
8. A must read: http://reports.weforum.org/future-of-jobs-2016/employment-trends/
9. Very informative data in education: https://ourworldindata.org/global-rise-of-education
11. Experience of autonomous driving in a Tesla: https://www.youtube.com/watch?v=3yCAZWdqX_Y
12. Hans Moravec, Robotics Institute, 1998 ftp://io.usp.br/los/IOF257/moravec.pdf
13. For further readings about the topic: http://futureoflife.org/background/benefits-risks-of-artificial-intelligence/