The text by prof. Dr. Maciej M. Sysło, describes the changes awaiting Polish education. The article also provides a hint and encouragement to the environment creating the Integrated Qualifications System - including AI companies involved in artificial intelligence - to work on describing qualifications to support teachers in adapting to the upcoming new school reality.
One of the missions of school is the intellectual development of students, who are now increasingly surrounded by devices and manifestations of artificial intelligence (AI). This happens when they program robots, play various games, use maps or Google translator. Outside of school, they also see robots replacing humans (e.g., on automobile assembly lines) and many other devices, machines and everyday objects exhibiting certain characteristics associated with human intelligence.
The main purpose of this chapter is to introduce AI to the educational community and to draw attention to and provide arguments for the fact that the time has come for artificial intelligence to appear overtly in schools. Because of AI’s close connection with computer methods and computer science, the computer science education community proposes that AI become a module of computer science education, as well as be integrated with other subjects.
A bird’s eye view of artificial intelligence… above the school
The traditional role of education is to prepare students for the future. In turn, the future today appears as constantly changing learning and work environments, which are increasingly filled with technology aimed at replacing routine work with automated procedures and devices. Over time, they are taking over many human functions traditionally understood as activities requiring intelligence. New professions are emerging that make increasing use of technology, including intelligent solutions, such as robots that automate more than just routine work. The ability to work with them will therefore be essential.
As has been the case with any previous technology, it is anticipated that AI can be an effective tool in education, relieving teachers of routine tasks (checking attendance, grading) and providing support in working with students. In turn, it will provide students with more effective and tailored learning methods. These could be interactive adaptive tutorials that create individualized learning environments and use a human-computer interface. They would monitor the student’s work, select appropriate tasks and provide full feedback to students. Combining such individual tutorials into a classroom system could be the beginning of a change in the role of teachers. However, it is important to keep in mind teachers’ perennial fears that any technology brought into the school could eventually diminish their role and in time eliminate them – they have examples of this in other automated workplaces. However, the role of teachers will have to be adapted to the environments in which AI-supported students will learn. So, as with other technologies, AI itself will require educational support. [...]
The objections to AI, in particular to machine learning (see subsection 4.3), prompt deep consideration of the place and way of using AI solutions in learning and in the education system. It seems that, at least at this stage, one should not completely surrender education to AI solutions, but rather adopt “mixed” solutions, representing the symbiosis proposed by Prensky of traditional thinking with artificial intelligence supported by machine learning. Artificial intelligence in education should therefore be looked at from the position of “artificial intelligence and the human” rather than “artificial intelligence versus the human”, that is, to consider how people can use AI for their own purposes: individually – in learning, as well as in education and in school – to organize the educational process.
AI is generally used in education in two areas:
AI in the classroom – the key issue is how to bring AI to students, what students should know about artificial intelligence, and how to use AI solutions in teaching and learning;
AI in the school – the transformation of the functioning of the school as an institution in terms of assessing the progress of students, organizing classes and the functioning of the school, as well as the form and mode of classes.
Artificial intelligence in the classroom
Education from an early age should train the ability to understand AI and communicate with systems using AI methods/algorithms. Properly reading and interpreting the effects of artificial intelligence is a prerequisite for the appropriate and safe use of its solutions. Unattended solutions (checkouts, cars) are already becoming commonplace today. The user should understand how they work, and thus have confidence in their safe and trouble-free operation. On the other hand, it is also important to learn how such systems should be and are built to better understand their operation. Many of these elements can already be conveyed to students at school.
In the near future, existing knowledge of the technology and its practical aspects will have to be augmented with basic knowledge of AI. Students should already start preparing at an early age to use AI solutions and navigate in an AI “driven” world, while being sensitized to ethical and social aspects. AI has the potential to change the forms and organization of teaching and learning. This will happen, for example, when applications for natural language recognition and processing begin to be used in language education. This expansion of artificial intelligence requires the inclusion of AI provisions in the computer science core curriculum. Its implementation should be accompanied in advance by providing schools with appropriate equipment and software, and especially by preparing and supporting teachers.
Selected proposals for lessons
Selected examples of the use of AI in various spheres of human activity are reviewed here. They can become the subject of activities to introduce AI and engage students to actively explore AI’s capabilities.
Robots are today the most popular educational toys, filling classrooms from kindergartens, through early childhood education (grades 1–3), to older grades almost to the end of secondary school. More advanced robot constructions are used and self-built in vocational schools, as an implementation of the provisions of the computer science core curriculum: “The student: [...] designs, creates and tests software that controls a robot or other object on a screen or in reality.” The robots used are equipped with a variety of sensors to recognize the environment, operate with sound and lights, and are generally programmable in various languages, such as Scratch, Blockly, Python. More advanced robots, such as humanoids, remain beyond the financial reach of the school.
An interesting story for students may be the history of attempts to automate the game of chess, which attracted some of the keenest mathematical minds. [...] Without going into the details of how the computer plays, you can offer students to play a game of chess against a computer at www.chess.com/pl. There are many other games available on the Internet that can be recommended to students as material for discussion of how much the computer in these games is exhibiting intelligence.
The computer translator is an excellent tool when using a foreign language, not only in the correct use of the language, but also in learning the ability to correct (debug) texts. Students can also use this program when working with source texts in other languages that they do not know.
Using Google translator can be a great opportunity, even a lesson, to make students aware that AI solutions can be imperfect in their intelligence. Every translation should be verified and often corrected. In general, this program has a problem with idioms, not to mention proverbs. It has already learned that “it’s raining cats and dogs” in Polish is “it’s pouring as if from buckets [leje jak z cebra],” but when asked to translate “raining cats and dogs” it gives us [the literal] “rain of cats and dogs” [deszcz kotów i psów], without guessing our mistake. Nor does it associate that “east or west home is best” in Polish means “everywhere is good, but home is the best” [wszędzie dobrze, ale w domu najlepiej]. However, it learns by accepting user corrections and suggestions. Translations are unlikely to be fully algorithmized, even by neural network methods, as these are sometimes unique. It needs to be trained on the translations of words or larger passages with cultural contexts.
There are many programs available that translate not only printed texts, but also handwriting, websites, spoken words and other forms of communication.
This is another service that uses machine learning. This application and others with similar purposes are the basis for planning and driving vehicles with autonomous driver functionality. We often use such systems to suggest our next moves on the road: go straight or turn right. In addition, they show the course of the entire route, the approximate travel time, some road signs (such as the speed limit).
[…] An interesting exercise with this system might be for students to conduct an experiment to verify the parameters of their various routes from home to school by different means of transportation: on foot, walking to pick up a friend, on a bicycle, by streetcar, by bus, by family car.
Students can already be partners in creating AI solutions using machine learning. The simplest examples of such possibilities are offered by the code.org environment. This is a collection of puzzles called Artificial Intelligence for Oceans https://code.org/oceans. In an interactive game, students teach an AI robot how to distinguish between fish and pollution so that the robot removes trash from the oceans. The puzzles are accompanied by videos from which students learn what AI is, what machine learning is, and what training data is. At the same time, they learn how artificial intelligence can be used to solve real world problems. These puzzles are accessible even to preschoolers. By solving them, children have the opportunity to take the first steps in learning about AI methods and applications.
Artificial intelligence in the hands of the teacher, in school
Given the ever-increasing presence of AI in our environment, we should define the place of AI in education and its contribution, as well as the role of traditional education in the environment filled by AI. It may also be useful to know the benefits of using AI in administering the education process. No less important are ethical issues, such as those relating to the purpose and scope of using the data collected by AI.
Old doubts resurface as to whether this time AI solutions will replace the teacher, if not fully, then to what possible extent? What then about the responsibility for the results of education, the competence of students, their education? Generally speaking, if in many areas AI is aiming to replace or at least displace humans, then education needs to rethink the competencies that humans should have in an environment with AI solutions.
As always, and this is repeated with every new technology or method of education, it is asserted that its goal is to improve the conditions and ways of learning and for students to achieve better results. Teachers, on the other hand, are told that the purpose of the technology is not to replace them. However, teachers are aware that every new technology means new responsibilities, not only relating to the need for further training, but also to daily duties in and out of the classroom. AI will be no exception here, in fact, it seems that it will require much more lesson preparation from teachers. And this doesn’t just apply to computer science teachers who will be introducing students to how AI works, but to all teachers when the classes they supervise/teach will be conducted in an environment of AI solutions. There are as yet no large-scale studies showing the benefits of AI-supported learning environments.
The greatest opportunities in AI applications in education are seen in adaptive (machine) learning systems, which adapt teaching and learning methods and materials to the capabilities and needs of each individual learner. […] Such a system can be used in the classroom, as well as in cases of individual tutors. […] Another popular area for such systems is foreign languages. A personal tutor is usually an open system, in the sense that provides the learner with opportunities for dialogue, questions and discussion, verification of solutions and dynamic assistance. Whenever such a system is used, it learns from data relating to the environment in which it is used.
Adaptive systems have many advantages, mainly for learners – they allow learners to have personalized, flexible and engaging learning. They can also support collaboration, group work, and are great for project-based work. As for teachers, such systems can be a catalyst for changing their role in teaching. However, it requires considerable preparation to use such systems. The teacher should know and understand the role of such a system, know how to interpret the data and results it provides, know how to formulate appropriate questions for the system, and work with the assistant of such a system as an integral part of it (intelligence). At present, the road to such systems in the classroom seems long: creating the system, preparing the teacher, teaching the system through the use of classroom data, and finally classroom activities. This path could be lengthened by the necessary expansion of AI environments in schools, which at this point seems unrealistic, although abandoning the classroom-teaching system and the school will do little to shorten the path. Real progress could be accelerated by the results of predictive studies that would define a roadmap for schools to gradually adopt AI technology.
No less important than the technical and logistical aspects are the ethical issues of using artificial intelligence in education. We may have doubts about the transparency of AI systems, understanding the decisions they make, their predictability, the possibility of auditing them, that is, controlling and supervising their operation. On the teacher’s side, on the other hand, is the responsibility for educational outcomes. What if such a system for some reason starts to malfunction due to errors in the “algorithm” or deliberate action? Such cases have already happened. One of the most important problems is also the protection of teachers’ and learners’ personal data and the right to access and use them. There are many other problems that AI will face in its expansion into the field of education, such as ensuring equal opportunities for all learners, in and out of school.
With the increasing impact of AI solutions on the daily lives of individuals, communities and entire societies, the ethical, legal and social issues are becoming more serious. Here we will comment on some media coverage regarding the impact of AI and how it is perceived.
On the first page of one of Google’s brochures (Grow with Google), you can read that “Google is committed to making sure everyone benefits from the opportunities created by new technology.” And on the last page of the brochure we read that “We are always inspired to see what people do when they have access to technology.” So Google is interested in watching us (to see what people do) and that inspires it to act (We are always inspired)! Those who are watched are generally unaware of this voyeurism (read – using the data they leave online, not always consciously) and do not know what the data will be used for.
One can question whether every AI solution is really a manifestation of intelligence. Is it a bot produced by a person who may not be very intelligent? Who is to judge this? A machine? The mere use of an intelligent tool does not guarantee the same in the final product. And with regard to education – how can the correctness of the methods used by an AI tutor be judged through the students’ performance? Or maybe the AI tutor will be judged by the AI tutor inspector?
Uber’s autonomous car has already killed a pedestrian in Arizona – the on-board system spent too much time analyzing the “object” it “saw” in front of it, instead of stopping immediately. Who is liable for such an accident? The owner (which could be the rental company), the passengers, the designer, or perhaps the programmer?
We will not continue to multiply examples to the detriment of AI. Fortunately, education has well-defined goals, the implementation of which is to point out to the student the benefits of AI in their near and more distant surroundings and the right ways of using it. It is also supposed to draw attention to the ethical and social aspects and risks. AI is expanding, perhaps even more so than in education, in almost every aspect of modern life. One can only hope that students’ attitudes formulated at school will direct their interests to the right track in using modern technology, which is increasingly “saturated” with AI solutions, and their human intelligence and sensitivity will prevail. […]
After taking into account AI's relationship with computer science, artificial intelligence should be integrated with other school subjects
Students now come into contact on an almost daily basis with devices whose results of the actions they undertake can be perceived as manifestations of intelligence.
[…] The computer science education community proposes that, after taking into account AI’s relationship with computers and computing, an appropriate AI module should be included in the scope of computer science education (computer science classes), in addition to being integrated with other subjects. The structure of such a module was taken from the vigorously developed US proposals, but its implementation will require taking into account the current core curriculum in computer science and other subjects. This proposal for curricular solutions is supplemented with examples of environments in which students can already develop familiarity with AI, as well as pay attention to important social aspects relating to the expansion of AI.
The article is a part of a broader publication on AI. You can download the entire publication using the link below (the text begins on page 74).
Author: Prof. Maciej M. Sysło – mathematician and computer scientist. In the mid-1960s, he watched as students from one of the high schools in Wroclaw ran their programs on an Elliott 803 machine as part of the country’s first classes in programming and numerical methods at school. Since the mid-1980s, he has been involved in preparing teachers for the expansion of computers in education. He considers his greatest success to be the maintenance, against trends at home and abroad, of separate computer science classes in schools, and more recently, the introduction of computer science education, including the teaching of computer science and programming to all students at all educational levels (primary and secondary). He is now popularizing computational thinking as a complement to the traditional 3R competencies of reading, writing and arithmetic. Honored with many national and international awards and grants.