After “Dancing with Robots”
The next government must have a vision of formulating a national education policy that can also address children living in marginalized, remote, disconnected and economically underdeveloped areas.
In the late 20th century, Frank Levy, Richard Murnane and David Autor looked into trends in the demand for skills by the United States workplace during the 1960-1990 period.
The study’s report has arguably been the 20th century's most influential writing on global education policy to date. In the report, they predicted that complex communication skills and expert thinking would increasingly be needed.
Complex communication is defined as a skill of interacting with other people to obtain information that can explain and convince. Expert thinking is described as expertise in solving a problem, for which a set of rules are not fixed in advance to do. These two main skills have become the basis for designing the Program for International Student Assessment (PISA) test instrument by the Organization for Economic Cooperation and Development (OECD).
Collaboration with machines
In 2013, Levy and Murnane expanded their study to cover the first decade of the 21st century, and the results of that study were published in their report Dancing with Robots. As they had expected, they found that the use of robots replaced the jobs of middle-payroll earners. Those jobs were mostly those that depended on routine and repetitive modes of fixed rules-based thinking.
In their 2012 reports, Henry Siu and Nir Jaimovich pointed out that the recovery efforts that were employed started in 2001 following two economic recessions and showed the largest employment growth had occurred in nonroutine-based work.
Will Indonesia experience such a phenomenon that has happened in the US? It is unlikely. A study is needed to examine this. However, employment anywhere moves forward in a generally similar nature. It is only the pace that may differ.
Also read:
> Us and Artificial Intelligence
Interestingly, Levy et. al also found that the types of jobs that incorporated human skills with machines were growing rapidly. It was the main message well represented in the title, Dancing with Robots, as mentioned above.
It is a dance that involves not only a combination of body and mind activities, but also an exercise of conscience. It means that humans must be ready to coexist and mingle with intelligent machines that might be programmed to have feelings like humans in the future. This type of human-machine collaboration is mostly what can be expected to emerge in the future.
Consequently, future education entails preparation and anticipation for the birth of various human-machine skills. For that, students need to get used to collaborating with machines. Education must systematically be designed for students to think and learn collaboratively with machines. The notion of complex communication should go beyond human interactions to communicate with machines.
Educators need to explore new learning opportunities based on human-machine collaboration. Building students learning capacity simply depends on a “copy-paste” of teachers’ knowledge should not be relevant any more.
Instead, the ability to formulate arguments and translate them as a series of instructions to solve a problem, so that it can be delegated to machines or a remotely working team, is as important as, or even more important than, the obtained solution itself.
This is in line with the fact that work can be transferred to another place easily with geographical locations no longer a constraint.
With the increasingly widespread practices of human-machine collaboration, it must also be admitted that various new problems have arisen. For example, with the increasingly effective functions of translating and grammar checking using artificial intelligence (AI), today many students write papers in Indonesian language and translate it into English with available automatic correction.
Some educators even recommend this method to the students assigned with the tasks under their supervision. This situation has forced educators and educational institutions to reflect on the main goals of education.
Also read:
> Digital Ecosystem not yet Established
In the field of mathematics, AI will force teachers to create many new questions. Trivial calculations, such as what is the mean of a large set of numbers, can be done within a click on computers. However, what if we are faced with the following problem: “Given four numbers with the smallest and largest numbers being eight and 21, respectively, is it possible that the mean is 19? Give an explanation.”
When I used language-processing tool Large Language Models (LLM), it replied, “There is no way the mean is 19. … if the smallest number is eight and the largest is 21, then the other two numbers are in that range as well. The sum of the four numbers would be between 32 and 84, and divided by four, it would give the mean between eight and 21, which is not equal to 19."
First, its argument is not appropriate because the mentioned sum of 32 or 82 is impossible given that two of the four numbers are already known. Second, its conclusion that 19 is impossible. Nor is it consistent with its preceding statement that the mean must be between eight and 21.
In the field of mathematics, AI will force teachers to create many new questions.
Then I input my response, "Are you sure?" It replied, “Yes, I apologize for the confusion I created. It is possible to derive the mean, if the other two numbers were 18 and 20, … the mean would be 19.”
Then I pointed out its mistake: “The mean would be 16.75, not 19.” He then apologized again, “I'm sorry, you're right. … I apologize for all the confusion caused by my previous statement.” Then, I asked: "So, what's the answer?" It replied the same as its first: "The answer is impossible, the mean is 19 ...". It ended up with an incorrect argument like the first one.
Actually, this problem indeed is a paradox, being "impossible" to get an answer, but the arguments it gave were not valid. In fact, the lowest mean could be derived if the other two numbers are equal to eight, and the highest mean is possible if the other two numbers are equal to 21. That is, the lowest mean is possible if the numbers are 8, 8, 8, 21 and it will be the highest mean when the data are 8, 21, 21, 21. So, the lowest mean is 11.25 and the highest is 17.75. So, it is exactly not 19, because it is greater than 17.75.
From this illustration, we hope that parents who still have children in school and math teachers will no longer worry about their children or students using AI. When students use AI and find themselves being stimulated to come up with arguments like the one above, they have the opportunity to learn to build rationality. They may now look at careers as lawyers or diplomats because they feel the pleasure of arguing based on rationality.
Likewise in language learning, AI must be used to refine the mind and hone the skills of its users. AI should not make it blunted. So, it is a misperception that use of AI in education is associated with plagiarism or cheating that may in some cases lead to use prohibition. The good or bad impact is a matter of how parents and teachers deal with it.
By “dancing with robots”, humans have found partners to debate and argue against.
Thus, by “dancing with robots”, humans have found partners to debate and argue against. In a few months, these partners will become smarter. Debating will shape up students’ reasoning capacity. How enlightening and refreshing this new educational practice is.
Seeing the potential for human-machine collaboration in this era, we need to redefine future education. Educational institutions and curriculum programmers must systematically incorporate machine collaborative capacity into learning programs.
Basic skills
Then, what about those currently pursuing their careers? Of course, they also need to enrich their working capacity by learning new skills. This is possible only if workers have built the capacity to further their skills, which Murnane and Levy refer to as “skills breeding skills”. This is a person who can master a new skill driven by their solid basic skills.
Mechanics, technicians, nurses or other professions assigned to new equipment need to read manuals that incorporate text and diagrams. Therefore, proficiency in reading textbooks and algorithms remains the main prerequisite for workers' career advancement. These basic skills should be honed from early ages. The learning process must start at home or in their environment.
Thus, children who grow up in well-educated and affluent families and environments, households with a tradition of reading from a lot of literary sources, benefit more. They are more likely to hone their reading skills than their friends from less educated and affluent families.
Economic inequality may breed gaping skills inequality and, in the future, it may even lead to increasing gaps in financial wellbeing.
This vicious circle must be broken with affirmative policies. The next government must have a vision of formulating a national education policy that can also address children living in marginalized, remote, disconnected and economically underdeveloped areas.
That every child anywhere has the right to learn high-quality basic skills must be the principle to drive national development. The indicator is clear that the quality of education a child deserves should not hinge on the zip code where he or she lives.
Iwan Pranoto, Mathematics lecturer, Bandung Institute of Technology (ITB)
This article was translated by Musthofid.
