The burgeoning field of artificial intelligence (AI) has led to the development of new educational approaches, particularly in the realm of gesture recognition and Internet of Things (IoT) device control. Despite these rapid advancements, practical applications and hands-on learning opportunities remain scarce. Many learners, including students, educators, and software engineers, have limited knowledge of hardware due to a lack of exposure to IoT, AI libraries, and human–machine interfaces. This gap is exacerbated by the absence of demonstrated examples and academic hardware journals. A significant challenge lies in the cumbersome process of updating IoT firmware, which is essential for incorporating new features. This paper introduces a novel solution that eliminates the need for firmware updates. By leveraging the Python Firmata library, applications on the host computer can be updated without affecting the IoT device’s firmware. The Firmata protocol enables seamless communication between the host and microcontroller, facilitating real-time interactions. Additionally, the abstraction capabilities of AI libraries, such as MediaPipe, simplify complex tasks into manageable components. For instance, MediaPipe provides precise hand landmark coordinates, enabling direct control of simple Arduino Nano devices without requiring detailed calculations. The paper’s contributions are valuable for a wide range of professionals, including mathematicians, AI engineers, software engineers, hardware engineers, IoT engineers, and network programmers.

Generative AI (GenAI) is rapidly transforming higher education. This study explores the imperative for curriculum reform to effectively integrate these powerful tools of GenAI into education and prepare students for an AI-driven world. It also proposes a comprehensive framework encompassing three key strategies: (1) fostering AI literacy across disciplines through tiered courses that address fundamental concepts, applied uses, and advanced techniques; (2) shifting pedagogical approaches from rote memorization to problem-solving, emphasizing active learning strategies such as problem-oriented and project-based learning, and encouraging interdisciplinary collaboration; and (3) establishing dynamic updating mechanisms of curriculum, including partnerships with industry and research institutions, modular curriculum design, and cultivating students’ self-learning abilities. Moreover, this study addresses critical considerations for successful implementation, such as faculty training, resource allocation, ethical implications, assessment strategies, and the maintenance of academic integrity in the face of AI-generated content. Furthermore, this study provides a roadmap for educators and institutions to navigate the opportunities and challenges of GenAI, empowering students to thrive in a rapidly evolving technological landscape.

This study examined the evolution of digital education policy in the United Kingdom (UK) from 2008 to 2024 based on a discourse analysis of 21 policy documents retrieved from the UK government’s official website, GOV.UK. For this purpose, the shift from early digital education initiatives to the integration of artificial intelligence (AI) in educational frameworks was traced and analyzed. The analysis revealed a progressive transition beginning with foundational digital literacy programs and infrastructure development and then followed by the incorporation of data-driven decision-making and personalized learning technologies. The most recent policies highlight the strategic adoption of AI to enhance educational outcomes, teacher support, and administrative efficiency. This evolution underscores the UK’s commitment to leveraging technological advancements to address educational challenges and prepare students for a rapidly changing digital economy. The findings provide valuable insights for policymakers and educators globally that emphasize the importance of adaptive policy frameworks that anticipate and respond to technological innovations.

Knowledge tracing (KT), aiming at mining students’ mastery of knowledge by their exercise records and predicting their performance on future test questions, is a critical task in educational assessment. While researchers achieve tremendous success with the rapid development of deep learning techniques, current KT tasks fall into the cracks from real-world teaching scenarios. Relying on extensive student data heavily and predicting numerical performances solely differ from the settings where teachers assess students’ knowledge state from limited practices and provide explanatory feedback. To fill this gap, this study explores a new task formulation, namely, explainable few-shot KT. By leveraging the powerful reasoning and generation abilities of large language models (LLMs), this study then proposes a cognition-guided framework that can track students’ knowledge from a few students’ records while providing natural language explanations. Experimental results from three widely used datasets show that LLMs can perform comparable or superior to competitive deep KT methods. Finally, this study discusses potential directions and calls for future improvements.

In recent years, significant advancements in educational technology, particularly the rise of online learning platforms, have transformed the way students engage with educational content. Most students currently access learning materials through live streaming or pre-recorded courses, but live sessions are often confronted with scheduling conflicts between instructors and students, thereby limiting their accessibility. This compels numerous students to opt for pre-recorded lessons, which, despite offering flexibility, lack real-time interaction and direct guidance. The unavailability of professional educators creates a challenge for parents, who are frequently unable to assist their children effectively with academic difficulties. Consequently, when students encounter obstacles during self-paced learning involving pre-recorded materials, their motivation to continue diminishes considerably. To address these challenges, this study developed an intelligent digital human-based teaching system that leverages the capabilities of large language models to provide personalized, interactive, and real-time support. It is designed to simulate the presence of a human instructor, enhancing the learning experience by offering tailored guidance, answering queries, and adapting to individual progress, features that ultimately foster a more engaging and efficient educational process. The proposed solution seeks to bridge the gap between the flexibility of pre-recorded lessons and the interactive, instantaneous engagement typically associated with live courses.

This paper explores the integration of virtual reality (VR) technology with molecular simulations to visualize structures, reactions, and behaviors across multiple scientific fields. An emerging tool, Manta, is introduced from its theoretical foundation, along with many immersive and interactive use cases. These virtual cases enable researchers and students to explore the structural properties of materials, to simulate dynamic molecular behaviors, and to observe complex chemical reactions in real time. In case studies of the microscopic mechanisms of aluminum/graphene, organic reactions like the Diels–Alder reaction, and the decomposition processes of energetic materials, Manta illustrates its ability to serve diverse needs. Further, the virtual docking experiment, which provides an ideal scenario for ligand–protein interactions, can be helpful for undergraduate students who are learning computational biology. Regarding computational chemistry, one of the cases describes the structure changes from a quantitative perspective, overcoming the steep learning curve associated with traditional methods. Overall, this paper highlights the potential of VR-enhanced molecular simulations to revolutionize scientific research and education, advancing the fields of molecular dynamics, materials science, and beyond.

This paper synthesizes four studies conducted at a special education independent school and affiliated liberal arts university with teachers, senior high school students, and college learners 18 and up, focusing on applying AI to (1) design course blueprints, (2) create comic strip assignments, (3) mediate interactive Socratic discussions, and (4) use learning data to assist students with disabilities in mathematics classes. Gordon Pask’s cybernetics is used to visualize interactions to show how AI acts as a component in emergent networks of minds in motion. The four sets of results, taken together, showcase how to implement principles of cybernetics in designing AI-mediated collaborative classrooms. Five out of six configurations of AI’s collaborative use outlined by Mike Sharples that the author’s research program has so far explored are presented through the four study scenarios and tied back to grey areas carved out by experts in AI education research concerned with design and implementation, classroom relationships, and assessment. Implications of current progress in the principal investigator’s research and further directions yet to be undertaken in implementing a series of subject-specific educational scenarios to utilize AI as a collaborative coach are discussed. Practical suggestions to shepherd effective AI-mediated curriculum design, classroom problem-solving and information acquisition, as well as nimble student evaluation are provided.