AI EdRevolution #11

Welcome to the 11th edition of AI EdRevolution, a guide to the intersection of AI and Education. This newsletter is published regularly, bringing the latest insights and developments in the field of AI in education. This edition explores the complexities of AI detection in education, ethical considerations, and innovative solutions to overcome these hurdles.

The featured highlight is an in-depth look at AI detection platforms and their challenges. As highlighted in articles from University World News, Times Higher Education, and Inside Higher Ed, AI detectors often fail to accurately identify machine-generated text, leading to false accusations and increased workload for educators.

AI detectors often fail to accurately identify machine-generated text, leading to false accusations and increased workload for educators

These tools can also discriminate against non-native English speakers, exacerbating educational inequalities. The newsletter explores how changing assessment methods and teaching students to use AI ethically can help mitigate these issues.

This edition also features a TEDx talk by Paul Matthews titled "AI in Schools: Cheater or Tutor?" which discusses AI's potential to enhance education through personalised support and teacher assistance. The talk provides insights into AI's dual role in maintaining academic integrity and supporting education. Additionally, the newsletter covers significant policy developments, such as Metro Nashville Schools' implementation of a new AI policy, addressing ethical considerations and privacy concerns in educational settings.

Emerging trends are explored through articles from arXiv and Higher Education, analysing research trends in AI education and reviewing empirical research on generative AI in higher education. The newsletter also touches on innovations like KlingAI, which leverages AI for video production in education, and discusses the growing demand for AI training among college graduates. Expert perspectives from various sources, including The Hechinger Report and The Statesman, provide a balanced view of AI's role in education, emphasising the importance of maintaining human elements in teaching while embracing technological advancements.

This edition aims to provide a comprehensive overview of the rapidly evolving landscape of AI in education, from classroom applications to policy frameworks, fostering critical discussions about the future of learning in an AI-driven world.

Quick Recap of Last Edition

The previous edition of "AI EdRevolution" explored groundbreaking developments in AI-enhanced education. OpenAI's ChatGPT EDU and Harvard Business School's AI-generated role-play scenarios highlighted the potential for personalised and engaging learning experiences. SimTheory.AI 's new Beta platform promised improved capabilities for existing educational AI tools, with a detailed review forthcoming.

Critical perspectives on AI in assessment were examined, with insights from Leon Furze emphasising the current limitations of AI in grading. The newsletter stressed the irreplaceable role of human intelligence in evaluation, noting that AI grading, while consistent, often diverges from teacher assessments and requires complex prompting.

The edition advocated for a shift from AI detection to teaching ethical AI use, promoting critical thinking and creative writing in a hybrid learning environment. It explored innovative tools like zero-shot essay scoring and the integration of AI writing assistants in classrooms, encouraging a balanced approach to AI adoption.

Policy considerations focused on developing students' ability to critically evaluate AI-generated content. The newsletter invited reader engagement on transforming assessments and enhancing education with AI, while maintaining ethical standards and fostering inclusive learning environments.

AI Detection Challenges in Education

There has been significant online discussion regarding the use of AI detection software platforms to determine the authenticity of student-submitted work, distinguishing between human-generated and AI-generated content. Currently, these platforms struggle to accurately detect AI-generated content, often misclassifying work created by AI as human-generated and vice versa. This inaccuracy poses serious challenges for educational institutions attempting to maintain academic integrity in an increasingly AI-integrated world.

False Accusations: One of the most significant concerns with AI detection tools is the risk of false accusations. As highlighted in the article from The Markup, AI detectors often fail to spot machine-generated text accurately, leading to students being wrongly accused of academic dishonesty. This issue is particularly severe for international students, who may be unfairly targeted due to linguistic differences. Additionally, Turnitin has the option to turn off AI detection systems, which can have severe implications for students. These false accusations can have devastating effects on students' academic records and mental well-being, as well as on the trust between students and educators. The Markup

GenAI Detection Tools, Adversarial Techniques, and Implications for Inclusivity in Higher Education This study investigates the efficacy of six major Generative AI (GenAI) text detectors when confronted with machine-generated content modified using techniques designed to evade detection. The results show that the detectors' already low accuracy rates (39.5%) drop significantly (17.4%) when faced with manipulated content. The study highlights the limitations and biases of these tools, particularly against non-native English speakers, who are more likely to be falsely accused of using AI due to their writing style. The authors emphasize that these tools cannot currently be recommended for determining academic integrity violations but may support student learning and uphold academic integrity when used non-punitively. The study concludes that a combined approach is necessary to address the challenges posed by GenAI in academia, promoting responsible and equitable use of these technologies Read More.

Increased Workload: The unreliability of current AI detection tools places an additional burden on educators. According to Martin O'Sullivan's who was interviewed on the Ai in Education podcast on using AI to detect cheating, AI detectors have falsely flagged students, causing significant frustration and stress. The workload created by these false positives includes the need for teachers to manually review and verify the AI's findings, address students' concerns, and maintain trust in the educational process.

AI Education Podcast

AI In Education

Researchers Deceive University Markers with AI-Generated Exam Papers: Researchers have successfully deceived university markers using AI-generated exam papers, showcasing the potential for artificial intelligence to mimic human-generated content. This experiment highlights the evolving capabilities of AI in replicating complex tasks traditionally performed by humans, raising questions about the future of assessment integrity in educational settings. The study underscores the need for vigilance and innovative strategies to detect and prevent AI-generated academic fraud, emphasizing the importance of maintaining the integrity of academic evaluations. This research highlights the urgent need for universities to reassess their evaluation methods and invest in more robust verification systems or change assessment methods to maintain academic integrity in an AI-driven world. Read More.

A Real-World Test of AI Infiltration in University Exams: A Turing Test Case Study This research article from PLOS ONE examines the potential impact of AI on the academic integrity of university assessments. Conducted by Peter Scarfe and colleagues, the study involved injecting 100% AI-written submissions into the examination system of a BSc Psychology degree at a UK university. The results were alarming: 94% of the AI submissions went undetected by exam markers, and these submissions received grades on average half a classification boundary higher than those of real students. The study highlights the substantial threat AI poses to academic integrity, emphasizing the need for universities to develop robust strategies to detect and manage AI-assisted academic misconduct. Read more: Read More.

Do Teachers Spot AI? Evaluating the Detectability of AI-Generated Texts Among Student Essays This study published in Computers and Education: Artificial Intelligence examines the ability of teachers to detect AI-generated texts among student essays. Conducted by Johanna Fleckenstein and colleagues, the research involved two experimental studies with novice and experienced teachers. The findings revealed that both groups struggled to accurately identify AI-generated texts. Novice teachers correctly identified only 45.1% of AI-generated texts, while experienced teachers performed slightly better, particularly with high-quality texts. However, teachers were generally overconfident in their judgments. The study highlights the challenge AI poses to academic integrity and underscores the need for robust strategies to manage AI-assisted academic misconduct. Read more: Read More.

Ethical Considerations and Solutions: The integration of AI in education brings about significant ethical considerations. The use of AI must be balanced with compassion and fairness, recognizing that students may turn to AI for various reasons, including time constraints, lack of confidence, or misunderstanding of academic expectations. Effective implementation requires proper training and support for educators, ensuring they understand these tools' workings, limitations, and result interpretation.

When addressing potential AI usage by students, Lance Eaton outlines how it's crucial to approach the situation with curiosity rather than accusation. Engaging in a conversation with the student to understand their process, challenges, and decision-making can provide valuable insights and learning opportunities for both the student and the educator. This approach allows for a deeper understanding of the student's needs and can help identify areas where additional support or guidance may be necessary.

The question of how students should use AI tools responsibly is crucial. As discussed in a Fair Observer article , AI grammar tools can help students improve their writing, but there's a fine line between using these tools as learning aids and becoming overly reliant on them. Educators need to guide students on how to use AI tools ethically, ensuring they enhance rather than replace the learning process.

Lance Eaton India Nye Wenner

While AI detection tools aim to maintain academic integrity, addressing their challenges is crucial for creating a fair and supportive educational landscape. Rethinking assessment methods and teaching ethical AI use can mitigate many issues associated with AI detection, fostering a more equitable learning environment.

To reduce reliance on AI detection, educators should consider diversifying assessment methods. This includes incorporating more project-based learning, oral presentations, and collaborative work. Simultaneously, integrating AI literacy into the curriculum can help students understand and leverage AI responsibly.

Ultimately, the goal is to ensure fairness and accuracy in AI-driven educational systems. This requires developing sophisticated, unbiased tools that maintain academic integrity and inclusivity. By fostering an environment of trust and open communication between educators and students, we can harness AI's potential while preserving the core values of education.

AI in The Classroom

AI in Schools: Cheater or Tutor? - TEDx Talk by Paul Matthews

This thought-provoking talk explores the dual nature of AI in education. Matthews discusses how AI can be both a potential tool for cheating and a powerful tutor, enhancing personalized learning experiences. He emphasizes the importance of guiding students to use AI as a supportive tool rather than a shortcut, highlighting its potential to provide tailored support and aid teachers in creating more engaging, adaptive learning environments. Matthews presents three key AI-enhanced strategies:

1. Text Differentiation: AI tools can adjust the readability of texts to suit different learning levels, ensuring all students can access and understand the material.
2. Retrieval Practice: AI can quickly generate quizzes, allowing students to practice recalling information without increasing teachers' workloads.
3. AI Feedback: Students can use AI to get immediate feedback on their work, helping them improve their own efforts rather than having AI do the work for them.

These strategies aim to empower both students and teachers in creating more effective, personalized learning experiences. Watch the video

Paul Matthews

Education Matters: Highlights the Use of AI in School Districts - New Jersey School Boards Association (July 2024)

This article provides insights into how school districts are integrating AI technologies to enhance educational outcomes. It covers various applications, from adaptive learning platforms to AI-assisted administrative tasks, showcasing how AI is reshaping the educational landscape at a district level. The piece also addresses the challenges and considerations schools face when implementing these technologies. Read more

Innovative Learning Through Gamification: Empowering Students to Create Science Games

This approach combines deep learning with practical application of AI in education. Students design their own educational games using AI, which requires a thorough understanding of the subject matter. Recently, Claude (Anthropic's AI) has included a new capability where students can write code using generative AI to create games in the Artifact section of the platform. This feature allows for the development of games that students can learn from, enhancing the educational experience. Challenges include ensuring students grasp scientific concepts deeply, integrating creativity with scientific accuracy, and refining games through continuous feedback. Solutions involve students diving deep into concepts before using AI to create games, utilizing AI for interactive iterations, and collaboratively enhancing games. An example game on photosynthesis demonstrates how students can apply complex scientific concepts in an engaging format. This method not only teaches science but also introduces students to coding and AI usage, preparing them for future technological landscapes. See My Example Here .

Should Students Rely on AI Grammar Tech to Write Better?

This article from Fair Observer discusses the use of AI grammar tools by students to enhance their writing. It examines both the benefits and drawbacks of relying on such technology. While AI grammar tools can help improve the clarity and correctness of student writing, the article warns that over-reliance may hinder the development of essential writing skills. Experts suggest that these tools should be used as supplementary aids rather than replacements for traditional writing instruction. The piece concludes that a balanced approach, integrating AI tools with strong foundational teaching, can best support student learning. Read more

Impact of AI-Based Tools on Student Performance: A PLOS ONE Study

A study published in PLOS ONE examines the effect of AI-based educational tools on student performance. The study reveals significant findings: students using AI tools achieved higher grades, with improvements averaging nearly half a grade boundary. Specifically, AI submissions attained grades at the higher end of the distribution of real student grades. The study also found that AI submissions had a significant grade advantage, with the most extreme differences approaching a full grade boundary. This research underscores AI's potential to enhance learning outcomes by offering personalized support and feedback that supplements traditional teaching methods. Furthermore, the study indicates that AI-generated answers were nearly undetectable, with 94% not being flagged by markers, posing a challenge for academic integrity. Read more

Voices from the Field: Expert Perspectives

Generative AI in Higher Education: A Systematic Review of Empirical Research

This systematic review examines the impact of generative AI on teaching, learning, and assessment practices in higher education. The study synthesizes findings from numerous empirical studies, highlighting the benefits of generative AI in creating personalized learning experiences, supporting academic research, and enhancing administrative efficiency. It also addresses the challenges and ethical considerations associated with the use of generative AI in higher education, providing a comprehensive overview of the current state of research and suggesting future directions for exploring the potential of generative AI in education. Read More.

VentureBeat - There's a Simple Answer to Address the Conundrum of AI Bias: More Diversity

This insightful article by Cindi Howson , Chief Data Strategy Officer at ThoughtSpot, argues that increasing diversity in AI development teams is crucial for creating fairer AI systems, including those used in education. Howson highlights the stark lack of diversity in STEM fields, with women making up less than a third of STEM workers and black professionals accounting for only 9% in math and computer science. She emphasises the need for early education and exposure to STEM for underrepresented groups, starting as early as elementary school. The article discusses how bias infects AI through biased training data and the personal biases of developers, citing examples from popular image generators. Howson stresses the importance of diverse representation in AI development to address these biases and create more accurate, inclusive models that benefit everyone in society, including in educational contexts. Read More

Policy Development

Georgia College Educators Advise State Senate on Educational Matters

Georgia college educators have been advising the state Senate on the integration of AI in education. Their discussions emphasize the importance of strategic planning, investment in AI infrastructure, and the development of policies that ensure ethical and effective use of AI tools. The article details specific recommendations made by educators, including the need for ongoing research into AI's impact on education and the creation of professional development programs to help teachers use AI effectively. These efforts aim to position Georgia as a leader in AI-enhanced education. Read More

Metro Nashville Schools to Implement New AI Policy

Metro Nashville Schools are set to implement a new AI policy that aims to regulate the use of AI in classrooms. The policy focuses on ensuring ethical practices, protecting student data privacy, and promoting equitable access to AI tools. It outlines guidelines for how AI can be used to support teaching and learning while preventing potential abuses. The article discusses the development process of the policy, including input from educators, parents, and AI experts, and highlights the anticipated benefits for students and teachers. Read More

Emerging Trends and Predictions

Artificial Intelligence in Higher Education: A Bibliometric Approach

This bibliometric analysis examines the evolution of AI research in higher education from 2000 to 2022. The study, conducted by researchers from the Central University of Kerala, analyzed 775 publications from the Scopus database. Key findings include a significant increase in publications over the years, with China and the United States leading in research output. The most frequently published journals were "Education and Information Technologies" and the "International Journal of Emerging Technologies in Learning." Emerging themes in AI research include machine learning, convolutional neural networks, curriculum development, and higher education systems. The study emphasizes the need for global collaboration to further advance AI technologies in higher education. Read More

OpenAI Co-Founder Andrej Karpathy Announces Eureka Labs, an AI Education Startup

OpenAI co-founder Andrej Karpathy has announced the launch of Eureka Labs, an AI education startup focused on developing innovative AI solutions for education. Eureka Labs aims to create AI tools that enhance learning experiences, streamline administrative tasks, and provide personalized support to students and teachers. The article provides an overview of Karpathy’s vision for the company, including its commitment to ethical AI development and its plans for collaboration with educational institutions. This new venture represents a significant step forward in the integration of AI in education. Read More

Dispatch from InstructureCon: All Eyes on AI & Lifelong Learning

InstructureCon 2024 showcased the latest innovations in AI and their applications in lifelong learning. Shiren Vijiasingam, Instructure's Chief Product Officer, discussed the company's focus on lifelong learning and the integration of AI in education. He described a tiered approach to AI adoption, starting with basic awareness and progressing to advanced integration and interoperability. Vijiasingam highlighted how AI tools can enhance teaching and learning by summarising large conversation threads, aiding in equity by translating languages, and providing actionable insights from data. Instructure's ongoing work includes developing standards for badging and comprehensive learning records, emphasising the importance of continuous learning beyond traditional degree programs. Read More

Derek Newton

AI Tools in Education

Kling AI: Revolutionising Classroom Learning with AI-Powered Tools

Kling AI is an innovative platform that uses AI to generate videos, transforming photos into dynamic videos or creating videos from prompts. This tool can be utilised by learners for projects and assessments, offering a creative and engaging way to present information. By converting static images into videos, KlingAI enhances the learning experience, making it more interactive and visually appealing. The platform's AI capabilities allow for easy and efficient video creation, helping students and educators alike to produce high-quality educational content. KlingAI's focus on accessibility and ease of use makes it a valuable resource for educators looking to integrate advanced technology into their teaching practices, I have attached a simple AI generated video of students on an excursion . Read more

Enjoy the newsletter? Share it with colleagues and friends and let's spread the word about the exciting developments in AI and education. Share your thoughts and ideas in response to this prompt. Let's explore the vast possibilities of AI together!

For detailed insights into these topics and to stay updated on the latest trends, tools, and discussions surrounding AI in education, stay tuned to AI EdRevolution.

I create these newsletters primarily for my own learning and understanding of AI in education. However, if others find them beneficial as well, that's an added bonus. The process of researching and writing these newsletters gives me the motivation to read more and, more importantly, to truly comprehend what I'm learning. I hope you enjoy reading this newsletter. Please let me know what you think - your feedback is valuable to me.

Other Articles and Guides By Miroslav Belik

Using AI for Practical Classes

Using AI for Marking

Using AI For Classroom Brainstorming

Using AI for Science Classes

Your feedback and suggestions are welcome for future issues. Connect with Miroslav Belik on LinkedIn

If any of these links are not working, please let me know, and I'll make immediate corrections. Your feedback helps me ensure the quality and accessibility of the information I share.