Artificial Intelligence in the Learning Domain!
AI is enabling machines to do tasks requiring human intelligence, machine learning is providing systems the ability to learn and improve automatically and researches in deep learning inspired by the structure, function and interconnections between billions of neurons is likely to lead to automation of many traditional roles and routine tasks. AI may bring newer opportunities and increased efficiency, but emotional intelligence is one area that machines find hard to emulate.
Technological advancement is reforming numerous functions, allowing one to accomplish more in less time. Google maps enabling us to reach new destinations with ease, text editors and autocorrect features in our devices, digital assistants, chatbots available 24/7 answering basic queries, search and recommendation algorithm’s saving our time, all seeking to enhance user experience.AI enabled online shopping giving us an amazing experience by offering recommendations matching our taste optimizing previous purchase data. Advanced software brilliantly assisting doctors with diagnosis and treatment options once the patient’s symptoms have been entered. When it comes to gathering data, analyzing it, interpreting and determining the recommended course of actions, machines can sometimes excel and do better than human beings. AI software development is a massive market that is continually upgrading.
The field of robotics is also growing rapidly. Humanoid robots are already being used in domains like research, space exploration, personal assistance, caregiving, education, entertainment, manufacturing, maintenance, public relations, and healthcare. In January 2020 ISRO unveiled Vyomitra, a female humanoid robot that can switch panel operations, environment control and life support systems, conversing with astronauts, recognizing them, and responding to their queries. The robot is said to be capable of doing multiple tasks. It is expected to fly in the first unmanned flight as part of the first human spaceflight programme (Gaganyaan), scheduled later this year. At CES 2020, Samsung’s Technology and Advanced Research Labs STAR unveiled Neon, a computationally created virtual being. Neon not only looks like a human but is capable of showing emotions, communicating with human affect, learning from experience and creating new memories. Not to forget mentioning Sophia the world’s first robot citizen and the first robot innovations ambassador for UNDP. Combining cutting-edge work in symbolic AI, neural networks, expert systems, machine perception, conversational natural language processing, adaptive motor control and cognitive architecture, Sophia can recognize human faces, see emotional expressions, and recognize various hand gestures. It can estimate other’s feelings during a conversation and try to find ways to achieve goals collaboratively. As per a report by Oxford Economics (2019) robots are expected to displace 20 million human workers worldwide by the year 2030.
The types of AI in operation based on functionality can be classified into reactive machines, machines with limited memory, a theory of mind and lastly consciousness.
Reactive machines are the most basic type of AI systems. They can’t form memories and can be used for automatically responding to a limited set or combination of inputs. Deep Blue, IBM’s chess-playing supercomputer is the perfect example of this type of machine which beat international grandmaster Garry Kasparov in the late 1990s. Deep Blue can make predictions about what moves might be next for it and its opponent, making the most intelligent move from the possible options.
The second class of AI, machines with limited memory like the self-driving cars can work using simple pieces of information from the past. However, such information is only transient created from the observations and added to the car’s pre-programmed representations of the world. So, in addition to having the capabilities of purely reactive machines, they are also capable of learning from historical data to make decisions.
Theory of mind is the next level of AI systems undergoing innovation where machines can form representations about the world and other entities to some extent. Theory of mind is the ability to predict the actions of self and others (Leslie, A.M,1987). The ability to read other’s mind is not only a distinguishing human quality but also a fine bridge between the existing machines and the desired ‘state of the art’ futuristic technology.
The actual development of the theory of mind generally follows an agreed-upon sequence of steps (Wellman, H. M. & Liu, D,2004). There are several developmental precursors that infants need like the concept of attention, understanding others’ intentions, ability to imitate, understanding of false beliefs and hidden feelings for the development of the theory of mind. Around the age of four, when children start to think about others’ thoughts and feelings, the true theory of mind emerges. Theory of mind help us communicate or offer our services considering the recipient needs. Relying on this with an understanding of self and other’s motives, we communicate, keeping in mind what others already know or what they don’t know. Theory of mind not only influence social interactions but is critical to how societies are formed and shaped.
The question that now arises is, whether the relationships between human beings and humanoid robots, can be characterized by a mode of interaction like the relationships between human beings that captures mental states and information both the expressed, implicit and unsaid? With the advancements in robotics attempting to create models with appropriate coordination of a large number of perceptual, sensory-motor, attentional, and cognitive processes; the challenge still is that how we build an artificial theory of mind in a robot akin to a human.
Machines with consciousness– Consciousness refers to an individual’s awareness of ones’ unique thoughts, memories, feelings, sensations making sense of one’s internal and external environment for a purposeful movement (Blackmore, 2004). This awareness of oneself is a subjective and unique experience. Building machines that can form representations about themselves and the world and who have this element of consciousness is being seen as a final step in artificial intelligence research. This type of AI will be capable of understanding and evoking emotions in others and also have feelings, needs, beliefs and desires of their own. Creating this kind of self-aware machines shall mark the pinnacle of success in AI research.
Appreciating the burgeoning discipline of artificial intelligence as discussed above, it seems intriguing to study how AI can influence the education sector and if humanoid robots can take up the role of teachers in the classroom?
According to a review published in the Science Robotics publication by Tony et al. (2018) “Robots can free up precious time for human teachers, allowing the teacher to focus on what people still do best: providing a comprehensive, empathic, and rewarding educational experience.” An example of it can be the robot Jill Watson brainchild of Professor Ashok Goel of Georgia Tech’s online Knowledge-Based Artificial Intelligence course. According to Goel, every time the course was offered, of the 300 or so students that enrolled they would post over 10,000 queries. In addition to being taxing on human teaching assistants, the questions were often repetitive. This led Goel to create Jill Watson, whose efficient handling of the student queries not only relieved them of excessive workload but positively influenced student satisfaction.
Artificial intelligence in education can deftly manage tasks such as taking attendance and routine grading. It could also help teachers in improvising course design promptly by generating new lesson plan suggestions and assisting in navigating online teaching resources. Not only this it can also assist teachers in giving them a more significant insight into student’s needs. Classrooms equipped with language processors, speech and gesture recognition technology, eye-tracking, and other physiological sensors can collect and analyze information about each student. The same information can then be utilized by the teachers to tailor their teaching strategies, matching it to student needs. An illustration of how modern technologies such as AI and robotics are being incorporated into some of China’s classrooms in efforts to lead the way in AI was published by Wall Street Journal (2019). Students at Jinhua Xiaoshun Primary School in eastern China begin their lessons by putting on headbands that uses three electrodes, one on the forehead and two behind the ears to detect electrical activity in the brain, sending the data to a teacher’s computer. The software generates real-time alerts about students’ attention levels and gives an analysis at the end of each class. A light at the front of the headband changes color to reflect a student’s concentration levels. Concerns are, however, being raised as to how well the technology can track concentration, the risk of false readings and the impact of AI surveillance on children’s mental well-being. Being constantly watched by an electronic eye in the classroom can have an impact on children’s psychological health, putting undue pressure on them. Worries are also being expressed by some parents on the possible misuse of student data making cyber security a key issue for consideration. AI is also being applied to education as a part of intelligent tutoring system (ITS) where a pedagogical agent can be designed to model interactions in the learning environment by assigning it different roles such as tutor or co-learner depending on the desired purpose of the agent (V.J. Shute, D. Zapata-Rivera, 2010).
Undoubtedly artificial intelligence and robots can complement and enrich the teaching experience for students and teachers when used wisely. The second part of the question whether humanoid robots can replace the human teachers in the classroom seems interesting as digital teachers can come with unique benefits. They won’t need holidays, monetary reinforcements or be late for work. A humanoid robot programmed for teaching a specific subject can be a vast reservoir of information as its system can be continually updated on the given subject from the beginning to the most recent literature. A human teacher’s knowledge is limited to their education or training and efforts made by them to stay updated with the latest developments in their respective field. Many times, owing to excessive teaching load, administrative responsibilities at work, family commitments or paucity of time and resources, it becomes challenging for them to invest in consistent efforts for self-growth or knowledge building.
However, despite the above benefits, an advanced humanoid robot taking up the role of a teacher in entirety seems not only a distant goal but not so desirable too. The role of a teacher is likely to remain irreplaceable by a machine for a long time to come for reasons like:
High dependence on social interaction and emotional intelligence skills: Teaching role is heavily dependent on active human interaction be it verbal or nonverbal communication. Humanoid robots may be efficient in sharing the relevant content and theories, but the way a teacher facilitates the class keeping in mind the nature of the subject, differing student abilities and create engagement is heavily dependent on human interaction and social intelligence skills. For instance, when it comes to teaching subjects like psychology, philosophy, literature and management, student learning is not confined to a textbook, course material and lectures. The free discussions, deliberations, debates, reflective exercises, carefully planned and executed group activities stimulating active participation from students, facilitated by a teacher all play a significant role in knowledge acquisition and skill-building. Such a participative experiential learning environment in higher education or adult learning leads to personal growth and learning of not only the students but the teacher also. However, it calls upon optimizing multiple skills beyond lecturing like empathy, patience, enthusiasm, self-management and playing a mediating role as a facilitator. Also, an able teacher modifies his teaching approach and methodology, observing the response and progress of pupils to achieve learning outcomes. So, the same teacher may adopt a different pace and teaching style for different students going by observational cues and learner’s preferences in the classroom. Considering these subtle aspects, behavioural skills, and high dependence on emotional intelligence, the success of robots taking up the role of a teacher in entirety is still a matter of doubt.
Vicarious learning: As per the social learning theory proposed by Albert Bandura (1965), much of learning takes place through observation. Observational learning can take place any time through environmental, social and cognitive interactions and influences. Students are more likely to imitate or learn behaviors from a teacher they perceive as a role model and who is a human like them rather than from a machine with differing constitution and abilities. For example, one extremely sincere university professor was left deeply displeased when someone entered late in class though he made it apparent only through facial expression. The environment of the university was relatively relaxed, and the professor could have adopted a similar attitude, but he was always on time for class. The teacher was greatly admired and loved by students for his excellent teaching style, disciplined attitude and earnest efforts he made for each session. Observing the example and standard set by the teacher for two consecutive semesters influenced the student’s attitude and behaviour for the better. Years later, the teacher is fondly remembered with gratitude by students not only for his teaching but for his integrity and for instilling the virtue of discipline in life. Vicarious learning is even more applicable in case of younger children who are keenly observing the actions of those around them, be it caregivers, teachers, siblings or friends. As technology continues to surface every sphere of our life, it will be worth reflecting in near future what kind of ‘models’ we want for our children- humans or humanoid robots and its implications too.
Inspire and motivate for high performance by communicating expectations: The Pygmalion effect, or Rosenthal effect, is a psychological phenomenon wherein high expectations lead to improved performance in a given area. Robert Rosenthal and Lenore Jacobson’s (1968) study showed that, if teachers were led to expect enhanced performance from children, then the children’s performance was enhanced. Lady Bird Johnson once said, “Children are apt to live up to what you believe of them’. Great teachers are able to appreciate students individuality, their performance in the development graph and encourage them by communicating realistic expectations.
If humanoid robots are well programmed to teach and grade student performance, they can as well be used to communicate expectations to the students. However, instilling faith in the student to strive for success in a given sphere in life requires an ability to see the student beyond his grades. Each student has unique aptitude, competencies, motivation, limitations and ambitions. The way a human teacher can understand, connect and motivate him to pursue in a given field communicating inside and outside the classroom engaging in formal and informal communication, can a humanoid robot replicate it with success?
Student engagement: Student engagement manifests in the classroom by the degree of attention, interest, participation, curiosity and involvement shown by students in the classroom. Meece et al. (1988) set a model for cognitive engagement seen in the learner’s participation, and interaction with the learning material, learning activities and learning community. In a world where information is available at a click of a button, teachers strive for developing and sustaining student engagement in multiple ways. They provide authentic and specific feedback, remove any barriers to learning, use creative methods to deliver content, clarify concerns, provide reinforcements, discuss real-life examples etc. Understanding the association between student engagement and performance, great teachers listen, tell stories, communicate to the best of their ability to stimulate student’s creativity. Enhancing student engagement requires the teacher to not only optimize on the knowledge of the subject but how that knowledge is transferred. Tony et al. (2018) in their article on social robots for education write that to build a fluent and contingent interaction between social robots and learners requires the seamless integration of a range of processes in artificial intelligence and robotics. Starting with the input to the system, the robot needs a sufficiently correct interpretation of the social environment for it to respond appropriately. This requires heavy reliance and development in speech recognition, visual social signal processing so that a robot can access the social environment. Speech recognition is still not so effective when it comes to understanding spoken utterances from young children and using touch screen to overcome this shortcoming can disturb the natural flow of the interaction. For robots to be autonomous, they must make decisions, take actions in a pedagogical environment with an understanding of the learner’s ability and progress. Building artificial social interaction like that between two humans is an essential prerequisite for enhancing student engagement that involves a wide range of cognitive and affective components, which is still a challenge in artificial intelligence and robotics.
AI is enabling machines to do tasks requiring human intelligence, machine learning is providing systems the ability to learn and improve automatically and researches in deep learning inspired by the structure, function and interconnections between billions of neurons is likely to lead to automation of many traditional roles and routine tasks. AI may bring newer opportunities and increased efficiency, but emotional intelligence is one area that machines find hard to emulate. According to a recent report published by Capgemini Research Institute (2019), Emotional Intelligence will be a must-have skill in the future, with its demand likely to rise six-fold within the next five years. The role of a teacher is not merely confined to sharing knowledge; but requires one to don multiple hats depending on the situation-be it a mentor, counselor, motivator, guide and ally drawing heavily on emotional intelligence skills adapting one’s approach as per the unique needs of different students. To conclude, artificial intelligence can be a very helpful assistant in making academic functions more robust, freeing up precious time for teachers and also complement the teacher in the classroom but taking up the role of the teacher completely seems inexpedient. In fact, seeing how the millennial and generation Z are absorbed in their electronic devices, the role of a human teacher will become all the more critical and ingenious.
- Bandura, A. (1965). Influence of models’ reinforcement contingencies on the acquisition of imitative responses. Journal of Personality and Social psychology, 1(6), 589-595
- Blackmore, S. (2004). Consciousness: An Introduction, 2nd Edn. Oxford: Oxford University Press.
- Bussing, K. (2016, September 10). Georgia Tech’s Teaching Assistant “Jill Watson” Turns Out To Be A Robot!. Retrieved 2020, November 11, https://bit.ly/3ngobJs
- Emotional Intelligence Report: Demand for emotional intelligence skills soars six-fold in response to the rise of AI and automation. (2019. October 17). Capgemini. Retrieved 2020, November 11. https://bit.ly/3ph2hHY
- How robots change the world. What automation really means for jobs and productivity. (June 2019). Oxford Economics. Retrieved 2020, November 11. https://bit.ly/3ljy88E
- Jones R.C (2019). Robots ‘to replace up to 20 million factory jobs’ by 2030. BBC News. Retrieved 2020, November 11. https://bbc.in/3eRxfBS
- Leslie, A. M. (1987). Pretense and representation: the origins of “theory of mind”. Psychological Review, 94 (4), 412–426.
- Meece, J. L, Blumenfeld, P. C, and Hoyle, R. H. (1988). Students’ goal orientations and cognitive engagement in classroom activities. Journal of Educational Psychology 80 (4), 514-523.
- Meet Vyomitra, first Indian woman to ride to space. (2020, January 22). The Hindu. Retrieved 2020, November 11. https://bit.ly/36opmju
- Rosenthal, R., & Jacobson, L. (1968). Pygmalion in the classroom. The urban review, 3 (1), 16-20
- Singhal N (2020). CES 2020: Samsung’s STAR Labs unveils ‘Artificial Human’ project Neon; shows human-like emotions, intelligence. Business Today. Retrieved 2020, November 11 https://bit.ly/3njPkLx
- Sophia, world’s first humanoid citizen, focuses on saving the planet, plans to conquer Mt Everest. (2018, March 22). The Economic Times. Retrieved 2020, November 11
- Tony B, J. Kennedy, A. Ramachandran, B.Scassella & F. Tanaka (2018) Social robots for education: A review. Science Robotics. Vol 3-21. https://bit.ly/3ndrMYS
- V.J. Shute, D. Zapata-Rivera (2010) in International Encyclopedia of Education (Third Edition) Elsevier Ltd.
- Wellman, H. M. & Liu, D. (2004). Scaling theory of mind tasks. Child Development, 75, 759-763.
- Yifan Wang, Shen Hong and Crystal Tai (2019). China’s Efforts to Lead the Way in AI Start in Its Classrooms. The Wall street journal. Retrieved 2020, November 11 https://on.wsj.com/32wuwso