What comes into your mind when you hear the word ‘Machine Learning’ and ‘Artificial Intelligence’? People commonly associate it with new tech and the fourth industrial revolution. But what is it exactly and why it matters? This commentary intends to bring your attention to the benefits that these two new techs provide in future education. Discussions about the changes in our education system seem to be delicate now and then, and this comes along with soaring fears on the massive replacements of human jobs through automation. The question of whether the adoption of new technologies would replace the human educators, therefore, becomes interesting to examine more than ever by academia and practitioners as it would also determine the efforts to engender quality education for the benefits of refining the quality of human resources.
It is desirable to say that future education embodies a system that could yield a greater opportunity for both the educators and the educated to engage in every teaching and learning activity, with an expected outcome of a sustained authenticity. With this in mind, one could agree that such aspiration would not be lived up unless there is a radical transformation to the kind of education we entertain today. As we follow, the conventional education purpose, which was to produce as many graduates as possible deemed ideal for production works, becomes more irrelevant to the era of disruption we live today at least due to the following reasons.
Future Education and the Skills for the Future Jobs
The World Economic Forum (WEF) listed a set of skills required to meet the demand in future workplaces. This includes complex problem solving, critical thinking, creativity, people management, coordinating with others, emotional intelligence, judgement and decision making, service orientation, negotiation and cognitive flexibility. These types of skills are named in order from the most in-demand one. As highlighted as well, skills needed to compete in the workforce in 2015 become less prioritized, i.e., negotiation and flexibility as machinery and automatization replace human's role in decision-making. In the similar vein, active learning would not at all be sought-after, as new tech like artificial intelligence substitutes human advisors for companies.
If we do not transform our education system, we would not be inclined to anticipate the fear of putting human labors at cost. In other words, our education practices should, of great importance, deliver such innovative methods that would equip us with the necessary skills for future jobs. This is not possible to be achieved with admirable ease, and the failure to address power struggles and competitions between institutions may hamper the realization. In a technical sense, there are some indicators we need to focus on improving. Factors such as educators, digital and hard infrastructures, as well as resources for education should also be ready to adapt to the new education system.
If we want to test the validity that the new tech could improve the human capital for future jobs by ensuring the quality of education, we might now look carefully at what could possibly be used and how it will deliver the quality of education. Machine learning, or “a field of computer science that uses statistical techniques to give computer systems the ability to “learn” (i.e., progressively improve performance on a specific task) with data, without being explicitly programmed” is deemed to provide benefits such as helping out human educators in many aspects: assessing students’ papers, predicting student performance, creating a customized system that would help reduce their difficulties and problems in school. The use of machine learning could ensure that the deliverability of future education would produce better outcomes. Along with the emergence of the new tech comes a feature we could enjoy: an adaptive learning theory, whereby there is a meeting point between the use of technology and the accommodation of specific and unique needs of each learner (all possible due to the customization and the tailor of education resources).
In the future education system, we could also establish very specific and detailed user profiles and personalization that together become crucial elements of a customized learning experience. Here in this regards, one might employ identification parts to be filled with user interests and learner profiles; and educational recommender would play roles in providing some background checks if the students feel more comfortable to consult real persons. Such things could be developed in many ways depending on the context of where the education process takes place. It would be aligned with the norms, cultures and environment a specific society has.
Now, the question is, what does it take to implement the use of machine learning in the education system? One might argue that it needs digital infrastructure, state policy, resources to finally adopt it. Policymakers should be aware of the use of this new tech and know how to bring about the best benefits it could offer. In this, collaborating with technocrats could help them understand it. Shall they put more lights on this matter of new tech and its utilization, and put efforts to consider its adoption into the education system deliberately, it is then reassuring to realize a betterment in human resources. Reforming curricula by incorporating some methods to employ basic programming and problem-solving would also be beneficial to equip the students with the future workplaces. Also, this could be maximized by the utilization of these new techs.
Once we employ the grading system with machine learning as the teacher’s assistance, people should no longer worry about the result of the grading. The chance for it to be invalid would be approaching zero as it takes a human to set the grading mechanism through coding, and there is a process of screening and assessment of a pool of various samples of students essays that later on would be established as a dataset. To accurately examine it, there are usually human checkers and machine learning. Standards would be established using the criteria in the dataset, and carefully examined one by one in the pool of essays. As a result, the grading system would likely to be less subjective as the machine can inquire why a paper deserves a better mark than the others; and this provides a chance for the two people involved in the grading process to cut the amount of time spent on grading by evaluating “language fluency, grammatical and semantical correctness, domain information content of the essays”.
Figure 1: Text mining in a train corpuses. Tested in the GMAT test.
Figure 2: Comparison of 8 essays from 8 sets graded by human expert and machine. Series 1 represents the human graded scores. Series 2 represents the machine-graded scores. Result: the machine is capable of assessing an essay like a human rater as the difference between both the scores is not much.
In this regards, machine learning provides us with a more objective way to evaluate a student’s performance as they do not possess emotions, intuition like humans do, even to break rules they think could benefit them. According to Steve Wheeler, “No chance of transgression of the routines entrusted to them to follow, nor to break a rule. They are unthinking, unfeeling and blindly loyal to the codes that regulate their function.” Computers are logical and follow instructions precisely, without deviation. The adoption of artificial intelligence and machine learning in the education system would redefine our education system, and hopefully would take e-learning to a brand new level, introducing a personalized and customized learning experience. The future education provides unlimited ways to make a great leap, one of which is by providing channels to create “Personal Learning Networks” (PLN), whereby the virtual and informal network of friends and resources to interact and to share info and knowledge according to the interests. The question that remains now is about the level to which we are ready to shift to this new type of education.
Unleashing the Potential, Breaking Down the Barriers
While it is true that the adoption of new tech is promising to leverage the quality of our education system, we still have some tasks in front of us to do. We should find effective ways to deconstruct the constructed knowledge whereby the Implementation of New Technology is solely understood in STEM-related studies. As this paper previously argues, machine learning and artificial intelligence could be used to grade qualitative essays. They could interpret the logic of thoughts and the structure of an article in a sophisticated manner and would result in an objective assessment. How possible is it to be brought into the context of Indonesian Education System? Can it possibly change our way of conducting the national examination, from the multiple choice one into an essay format? It resonates with the idea that we require the students to develop complex critical thinking to face the future of works.
Another thing we should bear in mind when we discuss this is the conundrum as to if human educators are replaced by machines, what tasks remain for them? Here, I contend that our task now is to prepare it in such way so that we could, rather than having to experience the loss of jobs due to the automation, use them in our best to support the roles of human educators. Together with this, we also should be ready to prepare the future consequences of the adoption of these new techs, as their effects could transcend into many dimensions of life, i.e., pedagogy, students, culture and broader society. To sum up, the role of a talented, inspirational, resourceful and skillful educator would not be replicated by other practice or approach. The most desirable approach, for now, is to systematically arrange a blended learning balancing system between the electronic and human contribution in the education system.
Editor: Treviliana Eka Putri
Read another article written by Kevin Iskandar Putra
 Gray, A. (2016). The 10 skills you need to thrive in the Fourth Industrial Revolution. [online] World Economic Forum. Available at: https://www.weforum.org/agenda/2016/01/the-10-skills-you-need-to-thrive-in-the-fourth-industrial-revolution/ [Accessed 15 Nov. 2018].
 Lynch, M. (2018). 8 Ways Machine Learning Will Improve Education - The Tech Edvocate. [online] The Tech Edvocate. Available at: https://www.thetechedvocate.org/8-ways-machine-learning-will-improve-education/ [Accessed 15 Nov. 2018].
 Montebello, M. (2018). AI Injected e-Learning: The Future of Online Education. 1st ed. Switzerland: Malta., p.73.
 Ramalingam, V., Pandian, A., Chetry, P. and Nigam, H. (2018). Automated Essay Grading using Machine Learning Algorithm. Journal of Physics: Conference Series, 1000 012030., p.3.
 Ramalingam, V., Pandian, A., Chetry, P. and Nigam, H., p.6.
 Montebello, M., p. vii.
 Montebello, M., p. 39.
 Montebello, M., p. viii.
 Montebello, M., p. 71.