In today’s society, technology appears to be taking over. Everywhere you look, people are glued to their phones, computers or televisions, where they are constantly processing information received from these devices. In some ways, this may be concerning to feel as though we are constantly consumed by our technological devices, but stop and consider how technology has helped vastly advance our society. The development of technology has provided us with tools that have changed the way we communicate, work and have fun, and we have just barely scraped the surface. One of the areas technology has been dramatically affecting is education. As explained in “Teaching in a Digital Age” (2019), throughout history the way education has been provided has been constantly adapting from reading books, to using the internet to now taking courses fully online. Technology has been able to provide students with new ways to learn that can even be tailored towards their specific needs, therefore making education more beneficial to students. 

One of the new technological advancements that is emerging is artificial intelligence (Bates, 2019). Artificial Intelligence is explained as “using computers to do things that traditionally require human intelligence” (Shroff, 2019). From experience, there are many people who believe that artificial intelligence are simply robots that learn from humans, but when researching the topic, artificial intelligence is so much more. Some examples of artificial intelligence that people use in their everyday lives are task automation, social media and self-driving cars (Shroff, 2019). As artificial intelligence continues to develop with other technological advancements, the opportunity for artificial intelligence to be used in many more aspects of life is inevitable. Specifically, we believe that artificial intelligence will have a large impact on education in the future. 

What is artificial intelligence?

In the article “Letting Artificial Intelligence in Education Out of the Box” (2016), Timms explains how artificial intelligence is currently used in the education system. As of today, artificial intelligence is used in classrooms mainly in the form of a computer, or as the author puts it a “‘black box’” (Timms, 2016). With the current use of artificial intelligence in education, teachers are left with a tool that is very limited to use within their classrooms (Timms, 2016). The reasoning behind this limited use to artificial intelligence in the classroom, may just be due to the fact that this technological advancement has just not yet been developed enough (Timms, 2016). According to the article, in the past 25 years artificial intelligence has made leaps and bounds, now becoming extremely useful in the business industry (Timms, 2016). If artificial intelligence has been able to make these steps into helping businesses in the past 25 years, what is stopping artificial intelligence from changing education as we know it? The article proposes two artificial intelligence options that could change the education field, being “robotics” and “‘smart classrooms’” (Timms, 2016). 

The first type of artificial intelligence proposed for education is robotics which has been a field looked at by scientists in developing “socially assistive robots” that are fully functioning and can assist people with everyday tasks (Timms, 2016). According to Timms, he believes that since these “socially assistive robots” have already been developed for specific tasks, creating educational artificial intelligence robots (what Timms calls “cobots”) is something that can be achieved within the next 25 years (Timms, 2016). The idea behind using “cobots,” is these robots would not take over the teachers job, but rather be used in addition to a teacher as “a precursor to learning” (Timms, 2016). Since teachers (for the most part) are the only individual in charge of teaching a classroom of students, having a robot to assist the teacher could be greatly beneficial to providing an improved open learning environment (Timms, 2016). The robots could have the capability to assist students with questions, “tailor” education to meet a students needs and even “flag” students that need extra assistance to provide this information back to the teachers (Timms, 2016). Using robots for education can provide a whole new educational environment that is student centred and flexible to each and every student’s needs. In the years to come, if robots were used in education classrooms would no longer have one teacher, but a teacher and robot with many different capabilities.

Robotics In Education:

 

To analyze the proposal of fully functioning artificially intelligent robots in education, we used the “SECTIONS Model” from the academic tool “Teaching in a Digital Age” (2019). The table below highlights how the use of an artificially intelligent robot could be beneficial as an open education resource and also how it could be a disadvantage.

SECTIONS Model	Advantages	Disadvantages
Students	-AI within a robot has the ability to be tailored to students of different backgrounds based on the design of the robot. -The system can be designed to be diverse and provide assistance based on students’ learning capabilities.	-Access to the robot may only be in a classroom. -Cost to using a robot is high, not all schools could afford to provide each class equal access.
Ease of Use	-The robot has complex functions to be developed, but once developed it can act as a teachers assistant and can communicate effectively with all users. -System can be designed specifically towards different environments, making the robot universal overall.	-Reliability of robot. -Requires outside maintenance for problems. -High costs to fix issues that arise with the robot.
Cost		-Costs thousands of dollars to purchase a single robot, making it unrealistic for any school to own 1 let alone a robot for each classroom. -Any issues require significant funds to be fixed. -Hiring a technological company for repairs is expensive. Having a technical analyst on campus to diagnose problems would add additional costs to each school.
Teaching and Media Selection	-Can be personalized based on the targeted learning environment. -Can be designed to provide modality through narration and communication with students. -Can have multimedia access by using a screen as a user interface to interact with the users. -Can be developed to have a voice for communication with the users. -Can provide information in segments to students based on the learning environment it is designed for. -Can be very interactive with users and mark issues with students to gain knowledge on each student. -Can be developed to ‘fit’ certain grade levels and keep information in understandable terms.	-Requires significant pre training for teachers to understand complexity of robot and use it effectively in the classroom.
Interaction	-Can be designed to verbally communicate with students and teachers to provide best learning experience for both parties. -Can have additional features such as a screen which displays information and reinforces through the verbal communication.
Organizational Issues	-Can be tailored to fit any classroom or environment, as the developers can create an AI for any setting.	-Cost is out of the majority of schools budgets. -Requires teachers to train and learn how to use the robot effectively.
Networking	-Can have an internet connection to use social media resources to enhance learning experiences. -Can access generated learning resources created previously by teachers. -Would be an open educational resource in the classroom as students can each use the robot for information.	-Access to robot from home would be limited (potentially access to robot through a computer).
Security and Privacy	-Can be built to ensure privacy of students and teachers. -Can have storage through a secure cloud or on a server for sensitive information. -Can be programmed to erase information once class has been completed.	-Could be considered invasive to students as the robot would need to record interactions with students and teachers.

The second type of artificial intelligence this article proposes could be used in education is the development of a “smart classroom” (Timms, 2016). To explain the concept of a “smart classroom,” Timms explains how in today’s society we have developed technology that provides us a “smart house” with the ability to turn on lights, control the temperature and monitor various aspects of our house (Timms, 2016). If scientists were able to take this concept of a “smart house” and transfer it to a classroom, there is the ability to provide a more engaging and supportive learning environment (Timms, 2016). An example of a tool that could be implemented in a “smart classroom” is the concept of a camera that uses facial recognition software (Timms, 2016). Facial recognition software has become a central piece in how we do things such as unlocking a phone, but Timms suggests that if we were able to use this same software in the classroom, we might be able to recognize a students facial expression when they are having difficulty with a task or to monitor “the emotional state of the learner” (Timms, 2016). By using a tool such as facial recognition, teachers could have the ability to know which students they may need to check in with to see how they are progressing and ensure that the student receives the assistance they need (Timms, 2016).This is simply just one of the proposed ideas from Timms for creating a “smart classroom” as the possibilities are endless and can even be combined with the use of an artificially intelligent robot (Timms, 2016). The idea of a “smart classroom” can use a variety of artificial intelligence in unison to create one cohesive “smart classroom” that provides students with the ability to receive the best education possible (Timms, 2016).

Smart Classrooms:

 

In the table below we have again used the “SECTIONS Mode” to analyze the possibility of using an artificially intelligent classroom in education in the future (Bates, 2019). The table highlights the key advantages and disadvantages of how a “smart classroom” could benefit education for students as well as teachers (Timms, 2016).

SECTIONS Model	Advantages	Disadvantages
Students	– By nature, a sophisticated machine learning algorithm would have the ability to adapt itself to each student’s unique backgrounds and learning needs. – Equipment would be installed in the students’ regular classroom, so access would not be prohibitive. – If used effectively, it will find the most preferred way of learning for each individual student.	– At this point, only primitive AI is prevalent in education, which would not be as effective at distinguishing between unique student’s needs. – Possibly dangerous to assume that students would be comfortable with this technology or any of its applications.
Ease of Use	– Students and instructors shouldn’t have to learn artificial intelligence, as it is just a process happening in the background of their chosen tool.	– Not so much a disadvantage of AI per se, but the application of AI and choice of ‘vessel’ must be efficient and easy to use, or else the AI risks being useless. – Reliability of the chosen algorithm is of utmost importance. If the AI is not thoroughly tested before release or has problems upon release, then the effect on student’s learning could have severe implications.
Cost	– As consumer AI has become more prevalent (smartphones, ‘smart home’ devices, etc), cost has decreased to reasonable levels.  – Maintenance cost (of the AI itself) should, by nature, be low if it is self-learning.	– Depending on the devices and functions chosen for the classroom, initial costs could be high. – Constructing a new instruction plan around new tools will initially be time-consuming. – Maintenance costs of the actual devices would not be predictable. – Often, AI and smart devices  are developed by a third party — this may involves licensing fees.
Teaching and Media Selection	– Smart classrooms (and AI in general) can be configured in any desired way; it is an adaptable system, so the instructor can plan any combination of approaches that meets the 12 principles of media design.
Interaction	– Sophisticated AI that is capable of providing feedback could essentially take the place of a human instructor when necessary – Well designed smart devices are capable of providing simple feedback for certain activities (see block device example from Letting Artificial Intelligence in Education Out of the Box). – The technology has the advantage of being able to be specifically designed from the ground up for interactivity.	– If sub-optimal devices are being used, the level of interaction and feedback needed by the student could be lacking.
Organizational Issues		– Navigating politics and bureaucracy in the attempt to implement new technology like this could prove to be complex; lots of variables that need to be agreed upon (cost, appropriateness, effectiveness, privacy/data collection issues, etc). – New support staff positions may be needed to maintain the devices and accompanying software. – It needs to be ensured that the instructors both understand and support the technology in order for it to be used effectively.
Networking	– Smart devices would have the ability to record a particular students performance and export/share the data to the desired platform.
Security and Privacy		– Privacy could be a significant issue with educational smart devices, just as it is with any other consumer smart device. If the AI and devices are coming from a third party, license agreements could dictate undesirable amounts of data collection from both the instructors and the students. – Data storage (even just temporary) becomes an issue; any data that needs to be stored would ideally be within the country that the devices are being used in.

We are currently witnessing artificial intelligence flourish within a variety of industries and consumer products, and are beginning to become aware of how powerful a tool it can be. Artificial intelligence and its potential applications show so much promise in education, but there are still obvious barriers that need to be overcome. They are not free from the privacy and security issues that current developers face constant scrutiny over, and costs to develop an efficient and sophisticated AI specifically for education would in the present day be quite substantial (although, as most technologies do, it is projected to get cheaper as time goes on). Reliability in particular, is a significant component that is absolutely vital to perfect before taking products and features like this into real educational environments. Students absolutely can not be used as ‘guinea pigs’ for any new educational tool — the tools, environment, and curriculum needs to work for them right from the start, or else their learning experience will be significantly compromised. 

Despite all of this, there are huge potential benefits to AI and its many forms if it’s implementation is executed correctly. Due to the self-improving abilities of a machine learning algorithm, there is almost limitless potential to what it can learn, understand, and then apply to the user of the system. The pertinent data it collects can be used in so many ways to improve the learning environment not only for the person using it in one environment, but for other people using the same AI elsewhere. Even interactivity, which one might think would be an issue when there is no ‘instructor’ directly involved with the interaction, is an aspect that can be overcome with a well-developed device. Out of our two different applications of AI, we feel that variations of the “smart classroom” are more realistic at the current point in time (Timms, 2016). Even though robots and robot technology have evolved significantly in recent years, it seems like it’s still a while off before a sophisticated robot that is capable of being an effective educational tool for a wide variety of students can be developed. With “smart classrooms,” you are just taking prominent consumer devices that are widely used (cameras, motion sensors, etc.), and integrating them within an educational context (Timms, 2016). This is a highly adaptable solution, and allows for a significant amount of configurability and customization toward the desired learning environment. With this sort of application, the AI working behind the scenes just needs to be developed to support student learning — by no means is this a small order, but it is more achievable right now than developing a fully functioning educational robot. It will be incredibly exciting to see how these devices, and artificial intelligence as a whole, gets applied to education!

Authors:

Joey Wingham & Ethan Waitzner

References:

Bates, A.W. T. (2019). Teaching in a Digital Age. Retrieved from https://pressbooks.bccampus.ca/teachinginadigitalagev2/

Shroff, R. (2019). Artificial Intelligence Explained in Simple Terms. Retrieved from https://medium.com/mytake/artificial-intelligence-explained-in-simple-english-part-1-2-1b28c1f762cf

Timms, M. J. (2016). Letting Artificial Intelligence in Education Out of the Box: Educational Cobots and Smart Classrooms. Retrieved from https://link-springer-com.ezproxy.library.uvic.ca/article/10.1007/s40593-016-0095-y