Generative AI Applications in Open Education: Opportunities, Challenges, and Scenarios

Whether it is the steam engine, electricity, computers, or the birth of the Internet, each has driven transformative social development. Every technological innovation brings about a qualitative change in life, work, and learning. The advent of the Industry 4.0 era has seen various new technologies accelerating a new round of transformations in education at an unprecedented speed. At the end of 2022, the American AI research laboratory OpenAI released the generative interactive tool ChatGPT (Chat Generative Pre-trained Transformer), igniting a wave of enthusiasm among users across the internet, with application scenarios emerging one after another. As global public opinion varies, how can we rationally understand the impact of generative AI on education? The report of the 20th National Congress of the Communist Party of China pointed out: “Promote the digitalization of education and build a learning-oriented society and country for lifelong learning.” Open education, as an important support for building a learning-oriented society, has long been closely linked to technological development. Whether it is television teaching, online teaching, or blended teaching, it cannot be separated from the empowerment of technology. Technology is neutral; what does generative AI mean for open education?

Overview of Generative AI

In 1956, the concept of “artificial intelligence” was first proposed by John McCarthy and others at the Dartmouth Conference in the United States. Since then, the connotation of artificial intelligence has deepened and expanded continuously. Bill Gates believes that “artificial intelligence is as revolutionary as the mobile phone and the Internet.” There are three typical classifications of artificial intelligence: weak AI, strong AI, and super AI. Generative artificial intelligence refers to the automatic generation of various types of content such as text, images, videos, and audio through AI-related technologies. Generative AI is experiencing explosive development and will accelerate the in-depth development of artificial intelligence. Foreign products of generative AI include: ChatGPT, Notion, Bard, Bing Chat, Chatsonic, Perplexity, etc. Domestic products in beta testing include: the Fudan University team’s ChatGPT-like model MOSS, Baidu’s “Wenxin Yiyan,” iFLYTEK’s “Xunfei Spark,” and Mita Technology’s “Dialogue Writing Cat,” among others, with more research and development on the way. Products similar to ChatGPT may be a turning point for strong AI, making AI that everyone can use and operate a reality.

The most typical product of generative AI is ChatGPT, which, as a chatbot program, can engage in high-quality, multi-turn conversations with people. It allows users to input text prompts and quickly generate text responses based on its knowledge gained through machine learning. It is a large language model (LLM) that can understand user intentions and generate corresponding replies. It can independently learn from data and produce complex and seemingly intelligent writing after being trained on a large dataset of text. On March 15, 2023, OpenAI released GPT-4, which has seen many new developments in image recognition capabilities, text input limitations, and answer accuracy. Shortly thereafter, OpenAI also released ChatGPT plugins, which means there are more possibilities for building a generative AI ecosystem. Professor Li Jiahou has given ChatGPT a Chinese name, “Liao Ge,” which vividly reflects its human-computer dialogue characteristics. Professor Gu Xiaoqing believes that ChatGPT will have a significant impact on the “what to teach, how to teach, and who teaches” in the education field, and it should be examined from the perspective of the ecological relationship between technology, society, and education, as it is becoming a force driving educational reform.

With the changes in social scenarios and the deep integration of information technology and education, the role of teachers is also changing. In the new wave of educational informatization, whether teachers can maximize the positive effects of information technology and curb its potential negative effects is one of the key factors determining the development of educational informatization. As one of the main bodies of educational reform, teachers must be endowed with new role meanings. In the era of artificial intelligence, from the perspective of social roles, teachers should shift to being leaders in “co-teaching and co-learning” with students, self-improvers, human-computer collaborators, and resource integrators. Professor Li Jiahou believes that the relationship between ChatGPT and teachers represents a new type of dual teacher-student relationship. From the perspective of teachers’ professional value, the role of teachers should seek transformation while maintaining core elements, shifting from “solving doubts” to “seeking doubts,” from “teaching” to “establishing a career,” and from “conveying the way” to “exploring the way,” actively responding to the challenges posed by machines to the teacher’s role in the age of artificial intelligence. From the perspective of educational scenarios, the intelligent era’s “seamless connection” and “cross-border integration” of educational scenarios will break the limitations of traditional classrooms, impacting teachers’ pre-set educational models, and the frequency of uncertainties and unknown situations will increase. In the yet-to-be-explored educational contexts, teachers are required to leverage their professional wisdom and the ability to transform knowledge into wisdom to cope with the complexity, diversity, and unknowns in the educational process. ChatGPT possesses powerful information integration capabilities and natural language processing abilities, freeing people from repetitive tasks and enabling collaborative innovation in human-computer interaction. The effective use of generative AI can enhance teachers’ efficiency and reduce their workload, allowing AI to replace teachers in completing some repetitive tasks. In the face of technological advancements, how AI robots and teachers can better integrate and assist each other, whether “machine + human brain” can achieve “dual brain integration,” what teachers should teach students, and how teachers, who mainly use the internet to educate students remotely in open education, will continue to serve as guides, helpers, resource providers, and supervisors, are all becoming inevitable transformations in the roles of teachers in open education.

The educational reform of the new era pursues the construction of a high-quality education system that is oriented towards and suitable for everyone. In 2022, “digitalization of education” was first included in the report of the 20th National Congress of the Communist Party of China. On February 27, 2023, the Central Committee of the Communist Party of China and the State Council issued the “Overall Layout Plan for Digital China Construction,” which pointed out that “the national education digitalization strategic action should be vigorously implemented.” The digital transformation of education has become a development trend for future educational innovation and reform, aiming to form a good educational ecology characterized by openness, adaptability, flexibility, and sustainability. Open education, as part of the national construction of a high-quality education system, has an inherent demand for the development of digital transformation in education. Artificial intelligence technology is becoming an important strategic support for the digital transformation of education. From the perspective of practical approaches, the integration of digital technology into educational fields will create a new semantics for the digital divide, fundamentally changing the practices of different educational subjects. From the perspective of application scenarios, artificial intelligence has enormous potential in educational management, learner assessment, teacher professional development, and other areas, playing a unique role in connecting fragmented decision-making processes and optimizing the supply chain of educational public services. Artificial intelligence will replace simple, repetitive cognitive labor, allowing teachers to fully leverage human advantages in innovation, complex decision-making, emotional care, and motivation. In the field of open education, the empowerment of artificial intelligence will connect open education, deepening its connotation, expanding functional requirements, and enhancing learning support, thus forming a pivotal force in the lifelong education system during the digital transformation process. Academician Zheng Nanning believes that ChatGPT has pressed the “accelerator button” on educational reform. Generative AI has the potential to promote the digital transformation of school education by reducing burdens and increasing efficiency, facilitating dialogue-based learning, serving as AI assistants and personal learning consultants, and personalizing educational instruction. Digital transformation is an important strategy for open education to respond to changes in the times and society, and how generative AI can assist in the digital transformation of open education and add momentum to its development is worth attention and consideration.

Competence refers to the individual characteristics that can significantly distinguish between high performers and average performers in a specific organizational environment, cultural atmosphere, and work activities. It includes motivations, attitudes, or values, as well as any knowledge or behavioral skills that can be measured and counted. For teachers, research includes existing teaching competence, digital competence, collaboration with families and communities in education, job competence, and professional competence. Teaching competence refers to the ability of teachers to effectively teach and continuously improve their subject-specific teaching levels and promote their professional development. In specific teaching contexts, the connotation of teaching competence varies. Teachers are one of the key factors determining teaching quality, and in the context of artificial intelligence, the teaching competence of teachers urgently needs to be enhanced. The intelligent teaching competence of teachers in this study refers to the personal characteristics, such as motivation, attitude, and philosophy, as well as the knowledge and skills required for intelligent teaching in the context of open education. Facing the rapid development of generative AI, teachers experience pressure from various impacts and challenges it brings, along with complex emotional responses. In the field of open education, teachers’ intelligent teaching competence will be an important factor in promoting the training of new higher education talents.

The primary audience of open education is adult learners, who generally possess certain experiences, have clear learning goals, and exhibit self-directed learning tendencies, often utilizing fragmented time for study. With the advancement of society, the learning motivation of adult learners has gradually shifted from meeting survival and career development needs to fulfilling self-value, quality development, and interest development needs, leading to a greater expectation for practical learning resources, diverse learning content, flexible learning methods, and learning environments. The emergence of generative AI will undoubtedly bring more demands and challenges to adult learners. In the age of artificial intelligence, information acquisition becomes more convenient, resources richer, and learning more flexible. Intelligent recommendations, assessments, companions, and tutoring will gradually become the norm, making it very common for learners to autonomously choose learning content, set their own learning pace, and engage in human-computer dialogue learning. The advent of ChatGPT has made student learning and teacher instruction more personalized and flexible, giving rise to new forms of educational instruction. In the new educational ecology under the background of generative AI, the requirements for students’ learning literacy are higher, urgently necessitating better self-regulation and management abilities, which will also lead to changes in students’ motivation to learn. Learning motivation is one of the important factors influencing the learning outcomes of adult learners. The integration of generative AI into open education will affect students’ learning motivation through factors such as cognitive interest, career development, breaking conventions, external influences, and social services, making it a challenge to leverage positive feedback for students.

Domestic scholars have likened generative AI to the “Aladdin’s lamp” and “Pandora’s box,” referring to its hidden potential and possible risks. ChatGPT can generate learning and teaching materials, intelligently answer questions, automate routine tasks, and assist in scientific research, but it also poses potential threats such as plagiarism, difficulty in distinguishing true from false information, privacy breaches, illegal behaviors from malicious use, ideological issues, specific value orientations, and social equality concerns. Although ChatGPT brings convenience to education, it is also a double-edged sword that may trigger a series of technological ethics, privacy protection, and digital divide issues in the education field. One of the greatest concerns among educators is that the popularity of ChatGPT may increase the likelihood of students using it to write assignments, significantly raising the probability of other academic misconduct. Some educational administrative departments and higher education institutions in the United States, France, and Australia have already ordered students to refrain from using generative AI for assignment writing. Generative AI may produce a lot of false information, potentially leading to serious public misguidance. Improper application of generative AI could create new trust crises. Even those with lower coding abilities can use it to create seemingly real news websites, where not only the news reports themselves but also the personal information of journalists and editors may be fabricated. Currently, generative AI is still immature; a technologically realistic attitude should be adopted, recognizing its limitations, while also valuing it but not relying on it. In the face of the potential negative impacts of generative AI, stakeholders in the field of open education need to enhance their risk awareness and remain highly vigilant regarding information security, ethical safety, and ideological risks to ensure the safe and controllable application of generative AI in open education.

Open education, in contrast to traditional closed education, features openness, lifelong learning, diversification, and respect for individual personality, consistently adhering to the basic principle of meeting citizens’ needs for open, lifelong, and personalized education, and always following the principles of “teaching without discrimination” and “teaching according to students’ aptitude.” As a more powerful artificial intelligence tool, the functions and application potential of ChatGPT are reflected in its ability to create customized learning plans and content according to students’ needs, automate repetitive tasks, create automated learning assessment systems, improve teacher efficiency, and generate various educational resources. Combining the characteristics of open education, adhering to the cultivation of virtue and fostering talent, and aiming to enhance teaching quality and meet learner needs, the following reflections on how to apply generative AI in open educational scenarios are proposed:

In open education, the vast majority of adult learners possess the ability to self-research and self-learn. Given the separation of time and space, information support is significant for adult learners’ studies. Generative AI can integrate information retrieval, intelligent services, chatbots, and creative tools, generating content that is suitable for user understanding and establishing relationships with users. The search interface of ChatGPT-like products is simple and can directly provide users with the desired information. Information retrieval powered by generative AI will be simpler and more effective. Teachers can utilize ChatGPT-like products for outline planning, data collection, and idea organization during lesson preparation. Students can use ChatGPT-like products for problem-solving, seeking guidance, obtaining learning materials, and receiving advice. Effectively integrating online courses in open education with generative AI can provide students with extremely simple and efficient information retrieval services, creating a new model of “searching is learning” through systematic course design and practice, utilizing derivative products with the characteristics of “minimal educational technology.”

For a long time, due to reasons such as geographical separation of teachers and students and conflicts between work and study, emotional exchanges between teachers and students, as well as among peers, have been insufficient or absent in open education. ChatGPT-like products possess a new mode of human-computer interaction, and their strong interactive capabilities mark the entry of the information society into a new era of human-computer interaction. In future teaching activities, the interaction between teachers, students, and machine systems will become more natural, providing a pathway for interactive learning. Dialogue-based learning supported by generative AI will become a new form of learning. For teachers, the application of generative AI will impact various aspects of teaching design, process, and evaluation; how to use it effectively will be a crucial consideration. A well-designed and utilized human-computer interaction will be an important lever for constructing a good learning ecology by effectively integrating teachers, learners, AI technology, and the teaching environment, providing more support and assistance for students in problem-solving, intelligent learning assistance, and emotional interaction, thereby generating new forms of interactive learning.

Open education learners exhibit significant age differences and diverse personalized needs, making tailored instruction challenging. Currently, open education predominantly adopts a blended teaching model of online and offline, with the collection, processing, and application of multi-dimensional data on learners’ behaviors, outcomes, assessments, diagnostics, and preferences gradually becoming normalized, leading to increasingly complete and accurate student digital profiles. A student digital profile is a collection of feature and behavioral labels processed and analyzed from various learning behavior and outcome data, which can assist teachers in timely adjusting teaching strategies and providing personalized services to meet learners’ needs. Personalized learning services aim to satisfy learners’ individualized, precise, and intelligent learning needs. In aspects such as intelligent academic diagnostics, learning path planning, learning resource recommendations, and precise individual guidance, personalized learning services can be generated, facilitating students’ personalized learning. On this basis, the effective integration of big data and generative AI can better understand students’ learning situations, interests, and needs based on their profiles, with ChatGPT-like tools serving as AI intelligent teachers providing more intelligent and personalized learning support services, helping students better acquire knowledge and skills. ChatGPT-like tools can become learners’ private intelligent assistants, capable of intelligently answering questions and providing various routine and personalized learning services.

Generative AI offers more imaginative and flexible space, compelling humans to rethink the function and value of the brain. Content-generating AI products create a composite brain for learners. Here, the composite brain refers to the internal brain of the individual connected to an external brain. The internal brain relies on the living organism itself, while the external brain is independent of the individual. By coupling with individuals, generative AI significantly enhances their perception and connectivity. For a long time, open education has emphasized evaluating learners’ outcomes, primarily combining formative assessments and summative examinations. Forms of formative assessment include scored assignments, written assignments, periodic tests, course practices, learning performances, online resource learning, group learning, reflective reports, etc. Generative AI products can solve many routine problems in learning. Once ChatGPT-like products are integrated into open education, students will develop a composite brain. Professor Zhu Zhiting believes that the existence of the composite brain may create a credibility crisis for teachers when evaluating students’ learning states. To prevent students from excessively relying on the external brain during formative assessments and neglecting the internal brain’s role, it is suggested that teachers reform the learning evaluation mechanism and establish a learning evaluation system based on the composite brain. For example, from the beginning of learning activity design, the application of ChatGPT-like tools can be embedded into learning activities, allowing learners to engage in collaborative learning and group learning based on generative AI tools, and evaluate the recorded learning processes. Students can also be assigned tasks based on ChatGPT-like tools, prompting them to comment, reflect, and improve upon their generated work. Especially for summative assessments in the form of papers and reports, teachers need to establish a new learning evaluation system from the perspective of the composite brain, reasonably evaluating the learning outcomes produced by the collaboration of the internal and external brains, stimulating learners to develop high-order thinking skills such as critical thinking and innovative thinking, and promoting learners’ discernment and selection of the answers or key points generated by the external brain, thus integrating the wisdom of the internal brain to achieve knowledge reconstruction and reproduction.

Conclusion

All that has passed is merely a prologue. Empowering education with artificial intelligence will be the vision for the high-quality development of future open education. The rise of generative AI, represented by ChatGPT, marks a significant transformation in the field of artificial intelligence, opening a new chapter in AI. In the face of a century of changes, education should focus on developing human qualities and cultivating the ability to actively adapt to the new era of artificial intelligence to meet future challenges. In the field of open education, the connection between technology and education has always been highly close, and the rise of generative AI will bring infinite opportunities and challenges. This technology is still in its developmental stage and has many shortcomings and limitations, but we should embrace intelligence with an open mind and practice with a positive attitude. The application of technology is not merely for the sake of using it but should be effective and purposeful. Only by adhering to a pragmatic and action-oriented attitude, starting from problem-solving and actual needs, and integrating technology with open educational scenarios can this new technology realize its due value, maximizing strengths while avoiding weaknesses, and utilizing resources effectively.