Introduction

The field of Artificial Intelligence in Education (AIED) has been around for decades. However, with the passage of time, it has gained more clarity as well as potential. AIED is high on research, policy, and political agenda around the world. The focus of the field remained, mainly, on the creation of systems and tools that can be as effective as one on one human tutoring or helping humans in tutoring on domain level knowledge. However, the focus is changing due to many factors. The use of AI, robotics, data science, and other related technologies would not only change the nature of many professions including education but also will change the types of skills valued in the market, generally referred to as 21st-century skills. The general learning skills and competencies such as critical thinking, metacognition, problem-solving, flexibility, and collaboration will be more important than technical and subject knowledge. New education theories also advocate for more personalization and agency in educational experiences. These trends show that not only AIED but education will go under a transformation in near future. The AIED will change to respond to the changing nature of education and education will also change due to the developments in AIED.
This paper is an attempt to provide a brief literature review of AIED with a special focus on its implications for teachers. The research questions to guide this study are: What role AI is expected to play in education and why? and what implications it will have on the role of teachers? The first section will discuss the role of AI in education in general and with respect to 21st-century skills. The next section will narrow down the discussion on the implications of AI on teachers. This review is based on the searches on the UMD library general search portal and DB Finder (EBSCOhost) under the education category, and Google Scholar. In addition to this general search, I also consulted the articles from the 19th and 20th AIED international conferences. I started with a general search of articles/books with keywords, “Artificial Intelligence in Education”, “Artificial Intelligence and Future of Education”, and Artificial Intelligence and Future of Teachers. I mainly focused on the latest articles in 2015 and onwards. I also narrow down my focus only to developed countries’ contexts. After an initial search of articles and books, I started reading the most relevant articles and used snowball techniques to find new articles (not limited to earlier stated timelines) from the above-cited sources. I included a few articles related to Robotics and 21st Century Skills in my search as I advanced on my topic.
A Case for AI in Education and 21st Century Skills

Two main reasons for the role of AI in education cited in most of the literature reviewed in the study are changing need of the market (Catalline, 1999; Hanna, 1998; McArthur et al., 2005; Lukin, 2018; Toumi, 2018; Selwyn, 2019) and advanced educational learning theories (Rolls & Wylie, 2016; Lukin, 2018; Robert-Mohoney et al., 2016; Selwyn, 2019). The availability of cheap processing power, big data availability, data storage capacity, and availability of fast internet have resulted in remarkable breakthroughs in the field of machine learning and artificial neural networks. Driverless cars, image analysis, real-time language processing, and translation, fraud detection, autonomous vehicles, automated customer service, synthetic art, process control, service robots, and many other applications are heavily relying on these two fields. Although the use of these applications is currently at a limited scale, these trends are an indication of a profound impact on the labor market and the competency requirements (Toumi, 2018). Given the value of the economy to a country, education systems tend to adapt to the current as well as futuristic needs of the market. In response to the AI age, some education functions will become obsolete and others will gain more prominence. The literature on 21st-century skills included in this review points out that there are many versions of these skills with some commonality as well as differences. The World Economic Forum in its 2015 report called New visions of Education: Unlocking the Potential of Technology identified the 16 most critical 21st-century skills required by future economies. The report divided these skills into three categories:
Foundational literacies that include conventional literacy, numeracy, ICT literacy, cultural and civic literacy, financial literacy, and scientific literacy.
Character qualities such as curiosity, persistence, adaptability, leadership, initiative, and cultural and social awareness.
Competencies comprise critical thinking, creativity, problem solving, and communication, and collaboration.
By contrast, Trilling and Fadel (2009) suggested a slightly different skill set again divided into three categories:
Learning and innovation skills: learning to learn and innovate, critical thinking and problem solving, creativity and innovation, communication, and collaboration.
Career and life skills: flexibility and adaptability, productivity and accountability, initiative and self-direction, social and cultural interaction, and leadership and responsibility.
Digital literacy skills: information literacy, media literacy, and ICT literacy.
These two examples provide a short overview of the debate concerning the skills required for the market in the future. Noah (2018) noted that current education systems are not designed to equip the students with most of these skills. ‘Numerous innovations in the education technology space are beginning to show potential in helping address skills gaps. These technologies have the potential to lower the cost and improve the quality of education’ (World Economic Forum, 2015, p-1).
The second type of impetus of AI in education comes from the new learning theories advocates for more student-centered, personalized educational experience, and agency (Collins & Halversion, 2010). The logic suggests that personalized technology represents a state-of-the-art alternative to traditional models of schooling, pedagogy, and organization that are irrelevant to the future digital age (Robert-Mohoney et al., 2016). Different components of artificial intelligence such as data mining, learning analytics, algorithm computations, and adaptive learning systems all have potential roles in this new emerging version of education. Regardless of the reason, the majority of the literature in the review agreed that AI will have a significant role in education. The role can be too broad to be covered in this paper. Therefore, I will limit my review on the AI implication on teachers only.
AI Implications for Teachers
There could hardly be any change in the education system that would not have implications for teachers. Teachers have been one of the main focuses on AI in education research and debate for many reasons ranging from cost to education quality factor. The articles reviewed in this study differ on what implications they foresee on teachers due to AI, other technologies, and 21st-century skills demands. Some argue for a central role of teachers with all decisions delegated to teacher and class level with the help of AI and robotics (World Economic Forum, 2015). A related view is of the shift of teachers’ role towards more complex learning tasks while routine tasks such as subject knowledge, literacy, and numeracy are delegated to AI (Luckin, 2018). Others pointed out the teacher’s role as a facilitator or data collector while classes are mainly managed by algorithms and data (Robert-Mohoney et al., 2016). Another view is the total replacement of teachers with humanoid robots (Selwyn, 2019; Timms, 2016). In the preceding section, the possible implications of robotics and AI on the role of teachers in the future will be discussed while the implications of other AI-related technologies are the topic of the last sections of the paper.
Robots in Education
Many industries around the world are facing increased high-tech automation. Industries such as underground mining, circuit-board manufacturing, and fruit-packaging rely on robots instead of human labor or humans working on machines. Intelligent systems are also expected to take over the place of humans in the profession like doctors, lawyers, and accountants (Noah, 2018). It has been generally assumed that education is one exceptional profession where the human teacher will remain central to education as learning is a process that requires social interaction with other humans with more knowledge and skills (Selwyn, 2019). There has been a significant advancement in the area of AI such as robotics and machine learning that present a prospect of changing the landscape of classrooms with numerous technologies and robotics in classrooms. ‘The introduction of robots could change education, especially to help prepare children with 21st-century skills and to increase student interest in robotics’ (Toh et. al., 2016, p- 151). The literature on robotics and education can be divided into two categories. One is about physical robots in the class, mostly at high schools and undergraduate level studies where students learn about robot functions and mechanisms while working on them. There is a push of introducing these physical robots related to learning in elementary schools as well. The other category is about far more sophisticated types of robots that are called ‘social robots’. These robots are designed for the classroom to act as teachers or companions.
The Japan Robotics Association (JRS), the United Nations Economic Commission (UNEC), and the International Federation of Robotics (IFR) report an increase in the market of social robots especially in the developed world (Pachidis et. al., 2019). In February 2010, the Ministry of Education in South Korea announced that they will equip robots for all 8,400 domestic kindergartens to facilitate instruction by 2013 (Wei & Hung, 2011). The empirical evidence of the impacts of robotics on student learning gain is limited but positive (Pachidis et. al., 2019). For instance, Wei and others (2011) did an experimental study based on the Joyful Classroom Learning System (JCLS). The study used robot learning companions (RLC) and other technologies. The aim of the study was to measure the impact of the experiment on 47 elementary (grade 2) students to learn mathematical multiplication. The result shows that the study design provides more hands-on exercises in a way that students have more thinking time for knowledge construction. RLC can increase students’ intrinsic motivations and offer a more joyful perception of the learning process. In addition, the JCLS provides a teacher with immediate and consistent information on the learning levels of every learner to adjust his or her teaching strategies in the classroom or after school support. However, it is not clear how much these results are specifically due to robots and not due to other technologies used in the design.
The literature concerning the impact of robots on student learning showed impressive results in language skill development (Toh et. al., 2016). For instance, Young and others (2010) used a robotic learning companion (a voice recognizing chicken robot) to see the impact on English language learning in Taiwan. They found the positive attitude of girls and boys students in grades 4 and 5 class towards robot learning companions. Such a setting motivates learners to engage in conversation without the fear of being right or wrong. They found students felt more happy and comfortable to engage with Rocky (the robot) than with the teacher. The robot design has limitations of the conversation only on a few topics but the results are promising. Similarly, Chang and others (2010) listed out the characteristics of a robot to be used for second language learning at the primary school level including repeatability, flexibility, digital data representation, humanoid appearance, body movement, interaction, and anthropomorphism. They also pointed out five modes of a robot that could be useful in language learning i.e. storytelling mode, oral reading mode, cheerleader mode, action command mode, and question and answer mode. Baker and Ansorge (2007) found that robots are effective at teaching 9-11 years students science, engineering, and other technical curriculums in terms of achievement score.

● There have been experiments with different types of robots such as Saya and Sota with different concepts. The result is mixed and the issue will remain as complex as robot sciences, AI, and learning sciences combined (Selwyn, 2019).
● However, the issue of the robots replacing teachers is not only the issue of feasibility but there are political, economic, and ideological factors that one should take into account while approaching these issues (Tuomi, 2018). AI in Classrooms for Teachers
● Most of the research in AIED has focused on the ways how teachers can use AI to
improve learning.
● Educational robots as Timms (2018) called them Cobot is also being worked upon as an assistant to the teacher.
● AI has the potential to assist the teacher in the form of software. For instance, it can help teachers in assessment in its various forms (Tuomi, 2018,  Nukwe et al., 2019), homework (Moore, et al., 2019), exam preparation (Le et al., 2018),
discipline (Selwyn, 2019), pedagogical assistance (Selwyn, 2019) and the list can go on.
● Al can not only help teachers to perform their job but it also has a strong application for preparing teachers for their job by supporting them in training and professional development (Porayska-Pomsta, 2016).

Conclusion
● Al will have a strong role in the future of education. Just like automation and AI will not make humans irrelevant but increase the value of certain skills while making absolute others, it will not replace teachers. However, the potential for teacher assistance seems countless.

References
Barker, B. S., & Ansorge, J. (2007). Robotics as means to increase achievement scores in an informal learning environment. Journal Research on Technology in Education, 39(3), 229-243.

Chang, C.-W., Lee, J.-H., Chao, P.-Y., Wang, C.-Y., & Chen, G.-D. (2010). Exploring the Possibility of Using Humanoid Robots as Instructional Tools for Teaching a Second Language in Primary School. Educational Technology & Society, 13 (2), 13–24.

Le, N., Shabas, A., & Niels, P. (2018). A Good Question Framework for Teachers. Artificial Intelligence in Education. 182-186.

Luckin, R. (2018). Machine Learning and Human Intelligence: The future of Education for the 21st Century. UCL IOE Press.

McArthur. D., Lewis, M., & Bishay, M. (2005). The Role of Artificial Intelligence in Education: Current Progress and Future Perspective. Journal of Educational Technology, 1(4), 42-80.
Moore R., Caines A., Rice A., Buttery P. (2019). Behavioural Cloning of Teachers for Automatic Homework Selection. In: Isotani S., Millán E., Ogan A., Hastings P., McLaren B., Luckin R. (eds) Artificial Intelligence in Education. 333-344

Ndukwe, I. G., Daniel, B. K., & Ahmedi, C. E. (2019). A Machine Learning Grading System using Chatbots. Artificial Intelligence in Education. 365-368.

Noah, Y. (2018). 21 Lessons for the 21st Century. Spiegel & Grau.

Porayska-Pomsta, K. (2016). AI as a Methodology for Supporting Educational Proxis and
Teacher Metacognition. International Artificial Intelligence in Education, 26, 679-700

Roll, I., & Wylie, R. (2016). Evolution and Revolution in Artificial Intelligence in Education. International Artificial Intelligence in Education, 26, 582-599.

Selwyn, N. (2019). Should Robots Replace Teachers? AI and the Future of Education. Polity.

Timms, J. M. (2016). Letting Artificial Intelligence in Education Out of Box: Education Cobots and Smart Classrooms. International Artificial Intelligence in Education, 26, 701-712.

Toh, L. P. E., Causo, A., Tzuo, P. W., Chen, I. M., & Yeo, S. H. (2016). A Review on the Use of Robots in Education and Young Children. Educational Technology & Society, 19 (2), 148–163.

Toumi, I. (2018). The Impact of Artificial Intelligence on Learning, Teaching, and Education. Policies for Future. European Union.

Pachidis, T., Vrochidou, E., Kaburlasos, V. G., Kostova, S., Bonković, M., & Papić, V. (2018). Social robotics in education: State-of-the-art and directions. In International Conference on Robotics in Alpe-Adria Danube Region (pp. 689-700). Springer, Cham.

Wei, C. W., & Hung, I., Lee, L., & Chen, N. (2011). A joyful classroom learning system with robot learning companion for children to learn mathematics multiplication. Turkish Online Journal of Educational Technology-TOJET, 10(2), 11-23.

Young, S. S. C., Wang, Y. H., & Jang, J. S. R. (2010). Exploring perceptions of integrating tangible learning companions in learning English conversation. British Journal of Educational Technology, 41(5), E78-E83.