PaperlayerView exact revision on arXiv

Source-linked AI summary

The AI Revolution in Education: Will AI Replace or Assist Teachers in Higher Education?

Cecilia Ka Yuk Chan, Louisa H. Y. Tsi

arXiv:2305.01185v1cs.CY

TL;DR

The study asks whether generative AI can replace teachers in higher education. Using survey data from Hong Kong university students and teachers, it finds that most participants view teachers as irreplaceable because of uniquely human qualities.

  • Problem

    The study examines whether generative AI can replace teachers amid concerns that AI may deliver standardized content and assessments more effectively.

  • Method

    The study surveyed students and teachers at Hong Kong universities about generative AI’s uses, perceptions, risks, and effects on teaching and learning.

  • Results

    Most participants considered teachers irreplaceable because of their critical thinking, creativity, emotions, and social-emotional competencies.

  • Takeaways & Limitations

    Teachers and universities should develop AI literacy and balance AI’s capabilities with human teachers’ distinctive qualities and human connection.

  • Takeaways & Limitations

    The comparatively small sample may not accurately represent all post-secondary educational institutions.

Abstract

from arXiv · show

This paper explores the potential of artificial intelligence (AI) in higher education, specifically its capacity to replace or assist human teachers. By reviewing relevant literature and analysing survey data from students and teachers, the study provides a comprehensive perspective on the future role of educators in the face of advancing AI technologies. Findings suggest that although some believe AI may eventually replace teachers, the majority of participants argue that human teachers possess unique qualities, such as critical thinking, creativity, and emotions, which make them irreplaceable. The study also emphasizes the importance of social-emotional competencies developed through human interactions, which AI technologies cannot currently replicate. The research proposes that teachers can effectively integrate AI to enhance teaching and learning without viewing it as a replacement. To do so, teachers need to understand how AI can work well with teachers and students while avoiding potential pitfalls, develop AI literacy, and address practical issues such as data protection, ethics, and privacy. The study reveals that students value and respect human teachers, even as AI becomes more prevalent in education. The study also introduces a roadmap for students, teachers, and universities. This roadmap serves as a valuable guide for refining teaching skills, fostering personal connections, and designing curriculums that effectively balance the strengths of human educators with AI technologies. The future of education lies in the synergy between human teachers and AI. By understanding and refining their unique qualities, teachers, students, and universities can effectively navigate the integration of AI, ensuring a well-rounded and impactful learning experience.

AI in Education · Literature on AI vs. Teachers

AI in education has evolved from computer-assisted instruction into diverse systems that personalize, automate, and immerse learning. Although AI may standardize delivery and support teachers, the literature emphasizes that human empathy, relationships, cultural sensitivity, and adaptability remain difficult to replace.

  • AI in Education: AI in education has developed from 1950s computer-assisted instruction into intelligent tutoring systems and virtual teaching assistants.The literature describes AI technologies ranging from 1-on-1 tutoring to virtual teaching assistance.
  • AI in Education: Dialogue-based and web-based AI systems can support knowledge co-creation, personalized feedback, and administrative work.These applications extend beyond intelligent tutoring systems into conversational learning support and task automation.
  • AI in Education: Humanoid robots, chatbots, and virtual reality can enhance engagement through interactive, personalized, and immersive learning environments.The literature identifies these as innovative technologies integrated into educational processes.
  • Literature on AI vs. Teachers: Some scholars argue that AI can deliver standardized content and assessments tirelessly and without bias, raising concerns about teacher replacement.This argument contrasts AI’s consistency and endurance with conventional teaching roles.
  • Literature on AI vs. Teachers: AI can efficiently assume administrative duties such as attendance checking, assignment and classroom monitoring, and paperwork.The literature presents task relief and improved efficiency as important benefits for teachers.
  • Literature on AI vs. Teachers: AI’s lack of sentience, self-awareness, emotions, empathy, and emotional intelligence limits its feasibility as a complete replacement for human teachers.The literature links these limitations to the social milieu formed by teacher-student, peer, family, community, and school relationships.
  • Literature on AI vs. Teachers: AI is characterized as a cognitive prosthesis that can aid teaching and learning but cannot yet replace human thought, values, or collaborative teacher-student relationships.This framing positions AI as support for education rather than a substitute for its human dimensions.
  • Literature on AI vs. Teachers: Human teachers contribute connection, cultural sensitivity, real-world context, curiosity, and adaptability that AI may struggle to replicate.These qualities support students’ personal growth, motivation, engagement with material, passion for learning, and responsiveness to individual needs.

Limitations of AI in Education · Literature on Collaboration between AI and Teachers

Research increasingly favors collaboration between AI and human teachers over replacement, with evidence that combining their strengths can improve learning and instructional performance. AI-supported applications can handle interactive, individualized, and administrative tasks while teachers retain instructional design and decision-making roles.

  • Literature on Collaboration between AI and Teachers: Researchers argue that collaboration between AI and teachers is more effective than treating them as opposing alternatives.Organizational research suggests that human–AI synergy can improve performance and compensate for each side’s weaknesses.
  • Literature on Collaboration between AI and Teachers: Combining human and machine intelligence has produced more effective learning outcomes than either working alone.Reviews also report increased teacher–AI collaboration in interactive learning environments over the past 20 years.
  • Literature on Collaboration between AI and Teachers: Jill Watson engaged students in online forums and answered questions about coursework and lesson content.Georgia Tech introduced this virtual teaching assistant in 2016 using IBM’s AI function.
  • Literature on Collaboration between AI and Teachers: In 2019, humanoid robot lecturer Yuki delivered preloaded lecture content and performed administrative duties in Germany.The example illustrates AI assisting lectures through prepared content delivery and support tasks.
  • Literature on Collaboration between AI and Teachers: Social robots provided verbal encouragement and gestures for remedial mathematics teaching, while conversational AI offered individualized language feedback and practice.These applications can address learning support gaps caused by teacher workload.
  • Literature on Collaboration between AI and Teachers: AI-generated individualized practice can allow teachers to focus more on instructional design and decision-making.Conversational AI addresses feedback and practice needs that may be limited by teacher workload.
  • Literature on Collaboration between AI and Teachers: AI can record students’ learning characteristics, analyze emotions, and use classroom monitoring technologies to support teaching and learning tasks.Examples include intelligent tutoring systems, sensors, monitors, and facial-recognition cameras.

Rationale for this study

This study examines whether generative AI can replace teachers and how it can work with or against them. It aims to help educators prepare for AI integration and identify ways human educators and AI can collaborate to enhance education.

  • The study asks whether generative AI technologies can replace teachers.
  • It examines how generative AI technologies can work with or against teachers.
  • Understanding AI’s potential impact on human teachers can help educators prepare for integrating AI into educational settings.The rationale emphasizes the need for educator readiness as AI continues to develop rapidly.
  • The study explores how AI and human educators can collaborate to enhance education rather than treating them as opposing alternatives.Educators are encouraged to optimize their value while co-existing and partnering with evolving AI technologies.

Methodology · Findings

The study surveyed Hong Kong university students and teachers about generative AI in teaching and learning, using closed- and open-ended questions. Data were collected through convenience sampling and analysed descriptively and thematically after pilot-based survey revisions.

  • Methodology: The survey examined generative AI usage and perceptions among students and teachers in Hong Kong universities.Questions addressed AI integration, including ChatGPT, associated risks, and effects on teaching and learning.
  • Methodology: The online questionnaire combined closed-ended and open-ended questions about AI technologies in higher education.The open-ended component supported thematic examination of participant responses.
  • Methodology: Participants were recruited through bulk email invitations using convenience sampling based on availability and willingness to participate.All participants received an informed consent form before completing the survey.
  • Methodology: The final sample included 384 undergraduate and postgraduate students and 144 teachers from various disciplines.The sample covered both student and teacher perspectives across disciplinary contexts.
  • Methodology: Descriptive analysis was used for the survey data, while thematic analysis examined responses to open-ended questions.The two approaches corresponded to the questionnaire’s quantitative and qualitative components.
  • Methodology: Two pilot-study rounds with 20 randomly chosen students and teachers preceded the main survey.The questionnaire was revised using pilot feedback and discussions with a research team.

Quantitative Data Findings

Survey findings show that students and teachers are open to generative AI’s educational benefits, but neither group strongly believes AI will replace teachers. Students generally reported more positive views than teachers regarding AI’s integration and benefits.

  • Survey design: The 11-item survey measured students’ and teachers’ perceptions of AI’s potential to replace teachers using participant counts, means, and standard deviations.Higher scores indicated greater agreement.
  • AI integration and benefits: Students reported greater openness to integrating generative AI into learning than teachers (M=3.86, SD=1.008 vs. M=3.61, SD=1.183; t=2.238, df=215.111, p=.026).The findings characterize students as more open to integrating generative AI technologies into their learning practices.
  • AI integration and benefits: Students rated AI more positively than teachers for improving academic performance (M=3.47, SD=.979 vs. M=3.29, SD=1.115; t=1.792, df=507, p=.074) and writing (M=3.32, SD=1.162 vs. M=3.01, SD=1.273; t=2.577, df=525, p=.010).Students were more optimistic about AI improving academic performance and helping them become better writers.
  • AI integration and benefits: Students rated AI’s ability to provide unique insights higher than teachers (M=3.74, SD=1.076 vs. M=3.47, SD=1.079; t=2.533, df=526, p=.012), while teachers more strongly believed students ask AI questions they would not ask human teachers (M=3.73, SD=.883 vs. M=3.39, SD=1.094; t=-3.695, df=308.968, p=<.001).Students also valued AI’s 24/7 availability (M=4.13, SD=.826).
  • Teacher replacement: Students and teachers similarly rejected the idea that AI will replace teachers (M=2.02, SD=.919 vs. M=2.03, SD=.946; t=-.057, df=539, p=.955).Both groups therefore did not strongly believe AI technologies will replace teachers in the future.

Qualitative Data Findings … (2) Replacing the Social-Emotional competencies developed from a Teacher’s Human

Qualitative findings show that most teachers and students do not foresee generative AI replacing teachers, although some consider replacement possible. Participants generally view human teachers as irreplaceable because AI cannot replicate human thinking, creativity, emotions, and contextual teaching input.

  • Generative AI technologies replacing teachers: Most teachers and students cannot foresee generative AI replacing teachers.Some students instead view AI as an auxiliary tool controlled by humans.
  • Generative AI technologies replacing teachers: Some students describe generative AI as a tool for asking questions rather than an autonomous teacher.This perception positions AI as supporting, rather than replacing, human educators.
  • (1) Replacing the Role of the Teacher: Some teachers and students nevertheless perceive that generative AI could replace teachers’ roles.Their reasoning emphasizes AI’s human-like communication and the accessibility of knowledge and data.
  • (1) Replacing the Role of the Teacher: Teachers may lose value if they continue using outdated teaching approaches while AI makes knowledge and data widely accessible.This concern links possible replacement to the continued use of old teaching methods.
  • (2) Replacing the Social-Emotional competencies developed from a Teacher’s Human: Participants identify human thinking, creativity, and emotions as qualities generative AI cannot replace.AI can gather and process information, but participants say it cannot expand or innovate like humans.

Generative AI technologies working with teachers · (1) Enhancing Course Planning, Design, and Pedagogy · (2) Developing Students’ Research and Writing Skills

Generative AI can collaborate effectively with teachers by improving teaching strategies, course design, assessment, and students’ research and writing. Its value lies in supporting educators and learners through practical, targeted educational tasks.

  • Generative AI technologies working with teachers: Generative AI technologies can collaborate effectively with teachers, optimizing teaching strategies and enriching learning processes and outcomes.
  • (1) Enhancing Course Planning, Design, and Pedagogy: Teachers use generative AI to brainstorm, summarize ideas, gather information, find knowledge, generate inspiration, and improve course planning, design, and pedagogy.
  • (1) Enhancing Course Planning, Design, and Pedagogy: Teachers design engaging courses with AI-generated scenarios, real-world problem-solving activities, and Kahoot multiple-choice questions tailored to students’ majors.
  • (1) Enhancing Course Planning, Design, and Pedagogy: Teachers use ChatGPT for assignment generation and critique, while students propose AI-assisted scoring and speech recognition to assess learning outcomes and oral presentations.
  • (2) Developing Students’ Research and Writing Skills: Teachers can use generative AI as a writing guide because it produces text with apparent structure, clarity, and logic, helping students improve writing.
  • (2) Developing Students’ Research and Writing Skills: Teachers assist students’ research by using generative AI to identify keywords, test search phrases, find references, and prepare academic paper reference sections.

(3) Preparing Students for an AI-driven Workplace and Future · (4) Improving Time Efficiency and Reducing Costs · (5) Encouraging Personalized Learning and Immediate Feedback

Teachers and students view AI as preparation for future workplace use and as a way to improve efficiency, reduce workload, and support personalized learning with immediate feedback. These applications position AI as a tool that complements teaching and learning activities.

  • (3) Preparing Students for an AI-driven Workplace and Future: Teachers emphasize that students must become proficient in using AI technologies and understand their implications for workplaces and future careers.Teachers describe AI proficiency as part of preparing students for lifelong access to these tools and for enhancing workplace productivity.
  • (3) Preparing Students for an AI-driven Workplace and Future: AI preparation is framed as a teacher responsibility because students will have access to these technologies throughout their lives.One teacher connected this responsibility to ensuring students can use AI effectively beyond education.
  • (4) Improving Time Efficiency and Reducing Costs: Teachers report that generative AI speeds routine tasks, accelerates lesson preparation, and supports assessment design, improving time efficiency and reducing costs.These uses also include administrative work, course logistics, tutorial registration, and generating email templates.
  • (4) Improving Time Efficiency and Reducing Costs: Students similarly perceive AI as reducing teachers’ workload, particularly for answering questions and generating lesson plans.This suggests that students recognize efficiency benefits for both instructional preparation and ongoing support.
  • (5) Encouraging Personalized Learning and Immediate Feedback: AI technologies can function as virtual tutors that provide personalized learning experiences and immediate feedback on student responses.ChatGPT is identified as capable of acting as a virtual intelligent tutor.
  • (5) Encouraging Personalized Learning and Immediate Feedback: Teachers envision generative AI supporting personalized learning by offering students advice or directions instead of directly displaying answers.This approach presents AI tutoring as guided assistance rather than answer provision.

Generative AI Technologies Working Against Teachers · (1) Lacking AI Literacy · (2) Failing to Ensure Equity, Preventing Academic Misconduct and Addressing

Generative AI may work against teachers when it undermines holistic competency development, reflects inadequate AI literacy, enables ethical or academic integrity violations, or encourages student over-reliance. The paper therefore emphasizes responsible training, governance, fair-use regulations, and safeguards for students’ originality and critical thinking.

  • Generative AI Technologies Working Against Teachers: Generative AI technologies may work against teachers in developing students’ holistic competencies.This concern is identified in the quantitative findings and reflects perceptions among some teachers and students.
  • (1) Lacking AI Literacy: Teachers warn that inadequate guidelines and training can cause generative AI to work against teachers.They advocate teaching students to use AI sensibly, responsibly, and professionally.
  • (1) Lacking AI Literacy: Insufficient AI literacy among staff and students is viewed as detrimental, especially regarding ethics, appropriate use, and equity.The proposed response is to develop AI literacy for both teachers and students.
  • (2) Failing to Ensure Equity, Preventing Academic Misconduct and Addressing: Governance of generative AI technologies is presented as necessary for addressing equity, academic misconduct, and responsible use.Teachers’ concerns include ethical violations, plagiarism, and potential loss of trust between students and teachers.
  • (2) Failing to Ensure Equity, Preventing Academic Misconduct and Addressing: Teachers call for regulations and frameworks to ensure the proper and fair use of generative AI technologies in universities.Although AI may benefit teaching and learning, teachers argue that regulatory mechanisms are needed for fair usage.
  • (2) Failing to Ensure Equity, Preventing Academic Misconduct and Addressing: Teachers caution that student over-reliance on generative AI can reduce original ideas and limit critical-thinking development.They associate over-reliance with laziness and a dearth of original ideas.

(3) Undermining Holistic Competency Development · Discussion

Generative AI may undermine students’ holistic competency development by limiting independent discovery, contextual understanding, and social-emotional growth. The study therefore frames AI as a teaching aid that requires human guidance, AI literacy, and continued recognition of teachers’ distinctive value.

  • (3) Undermining Holistic Competency Development: Generative AI may hinder students’ original discoveries because it can copy existing information instead of requiring independent research.Teachers also worry that AI-generated text may reflect little or incomplete understanding of context.
  • Discussion: Some teachers and students believe generative AI could eventually replace teachers, while others stress the importance of preserving holistic education.Holistic competencies include problem-solving, critical thinking, communication, and teamwork, which support students’ long-term personal and professional success.
  • Discussion: Teachers argue that generative AI cannot replace social-emotional competencies developed through interactions with human educators.They also identify cultural qualities and traditional values as capacities developed through human knowledge, experience, and interaction across diverse contexts.
  • Discussion: Rather than replacing teachers, generative AI can enhance teaching and learning when educators understand where collaboration works and which conditions make AI counterproductive.Effective integration requires attention to the dimensions of teacher–student–AI cooperation and the conditions that should be avoided.
  • Discussion: Students value and respect human teachers despite expectations that generative AI will become increasingly prevalent in education.The findings challenge the belief that younger learners necessarily prefer technology-centered learning.
  • Discussion: Teachers and students should harness AI through guidelines, training, and improved AI literacy so they can collaborate with it effectively.The study presents generative AI as a lasting technology with potential benefits for personal, social, and professional life.

Conclusions

The conclusions position human teachers as irreplaceable while urging teachers, students, and universities to integrate AI as a complementary tool. They propose refining human-centred skills, valuing personal connections, and designing education around a balanced human–AI synergy.

  • Implications: The study urges teachers and universities to reconsider what, how, and why students should learn alongside technology.The conclusions describe these questions as a “rude awakening” despite findings supporting teachers’ irreplaceable role.
  • Roadmap: Table 1 identifies unique human-teacher qualities across eight categories and 26 aspects, contrasting them with AI’s educational limitations.These qualities are intended to inform decisions about education’s future as generative AI becomes more influential.
  • Roadmap: Teachers should refine emotional intelligence, pedagogical skills, and personalized support while pursuing continuous professional development for effective AI integration.The roadmap focuses on abilities that are difficult for AI to replicate and on staying current with AI advances.
  • Roadmap: Students should value human teachers and seek personal connections because emotional and interpersonal skills support personal growth, resilience, and critical thinking.AI can provide resources and support but cannot fully replicate these human-centred learning opportunities.
  • Roadmap: Universities should design curricula that leverage AI to enhance learning while capitalizing on human-teacher strengths and accounting for costs, workload, and timing.AI is framed as a support tool rather than a replacement for human teachers.
  • Future direction: A symbiotic human–AI relationship can delegate mundane tasks to AI, preserve teachers’ personal focus, and require ethical advocacy in AI development.The conclusions emphasize designing AI to complement rather than replace human educators.

Limitations

The study is limited by its comparatively small sample, exclusive focus on text-based generative AI, and reliance on self-reported participant data.

  • Limitations: The comparatively small sample may not accurately represent all post-secondary educational institutions.This limits the study’s generalizability across higher education contexts.
  • Limitations: The investigation considered only text-based generative AI technology, excluding other forms or variations.This narrows the scope of the AI technologies examined.
  • Limitations: Reliance on self-reported participant data could introduce bias or inaccuracies.Participant reports may not fully reflect objective conditions or outcomes.
Loading 2305.01185v1…