"Hello everyone! I'm the digital assistant of Professor Liu Yanli at Tianjin University, your study buddy!"
On April 2nd, during the first class of "Fundamentals of Power Systems," undergraduate student Pan Linglan from the School of Electrical Automation and Information Engineering at Tianjin University was deeply attracted by another "Professor Liu" on the screen. She and her classmates whispered excitedly, full of curiosity about this new "digital teacher."
This "digital teacher" is the "AI teaching assistant for the whole process and all aspects," led by Professor Liu Yanli of the School of Electrical Automation and Information Engineering at Tianjin University. Behind this digital teacher, who simulates the image of Professor Liu Yanli at a ratio of 1:1, are even more powerful functions.
"The AI teaching assistant for the whole process and all aspects can cover various stages of the course and provide online services around the clock," Liu Yanli explained. "Before class, the AI assistant organizes the key and difficult points of the course knowledge, making it convenient for students to preview in advance. During the class, the AI assistant's Q&A system can solve students' course-related questions at any time, providing them with timely and convenient channels for getting answers. After class, the AI assistant provides test questions for each chapter's knowledge points, facilitating students in consolidating their knowledge. Additionally, it presents the overall knowledge system of the course through a knowledge map, making it convenient for students to identify and fill in gaps."
Of course, the capabilities of this "digital teacher" extend far beyond this.
The "Digital Teacher" Launches "Power System Fundamentals" Course Interviewee Supplied Image
Making "Tailored Teaching" a Reality
"In the past, teachers needed to personally record MOOC courses, and lesson preparation would consume a lot of time and energy," Liu Yanli explained. "However, the 'digital teacher' can generate MOOCs with just one click based on the teaching materials provided by the teacher, efficiently and promptly creating and updating MOOC resources anytime, anywhere, to systematically reinforce and expand students' learning content."
"After completing a class, if I don't have a deep enough understanding of the key points, I will watch the online course explanations in the AI assistant system and then do some test questions to deepen my understanding," Pan Linglan told the China Science Daily. Nowadays, AI assistants have become indispensable learning aids for students.
Another major system of the AI assistant, the "Question-Answering Mega Model," goes beyond classroom teaching to help students better accomplish advanced professional tasks.
Liu Yanli explained, "The 'Question-Answering Mega Model' is a domestically developed human-machine dialogue mega model system serving scientific research learning, similar to ChatGPT abroad." "On the one hand, it can provide 24-hour question-answering services, addressing students' questions and doubts in professional knowledge. On the other hand, similar to e-commerce platforms recommending personalized products to users, it can recommend different professional questions to different students based on their interactions, supplementing students' deficiencies in professional knowledge and achieving true tailored teaching."
In addition to basic question-answering, the "Question-Answering Mega Model" can also provide students with comprehensive support in programming skills. "The subject of power systems involves many analytical calculations that require programming to complete. The mega model can assist students in writing and debugging programs, saving time needed for programming skill training. This way, students can free up more energy to focus on advanced professional tasks."
This process is not only a professional practice but also a process of cultivating students' digital literacy. "Students only need basic programming thinking skills. By inputting instructions into the mega model, they can get the corresponding code. Now, students are very proficient in using the mega model, and they truly treat AI as 'productivity,'" Liu Yanli said proudly.
Design Concept Inspired by Virtual Anchors and Chatbots
"The integration of AI technology in the field of education is an inevitable trend," Liu Yanli said. With the rapid improvement of data storage and computing capabilities and the continuous breakthroughs in core algorithms, the intelligent teaching conditions supported by artificial intelligence technology will further expand the interactive space of teaching and promote the creative application of teaching methods.
However, many large models on the market still face risks regarding usage permissions. In this context, developing a domestically produced teaching tool serving students and creating a domestically produced universal scientific research and teaching platform has always been Liu Yanli's wish.
The inspiration for the "Digital Teacher" initially came from virtual anchors on the internet, Liu Yanli mentioned. "In the past two years, virtual anchors were frequently seen on platforms like Douyin, so I thought that applying this technology to the teaching scenes of higher education would significantly reduce the time and energy needed for teachers to record courses, allowing for more efficient provision of online learning resources for students."
In addition to course recording, personalized question-answering is also a much-needed function in the student learning process. Based on years of teaching experience, Liu Yanli thought that students' questions could basically be categorized within a certain range. "So, it is entirely possible to train a 'digital clone' of mine to answer these questions online 24/7. When human-machine dialogue technology emerged, I immediately had the idea of applying it to subject teaching."
In higher education, having AI assistants replace human teachers in the classroom requires AI assistants to possess extremely strong reserves of professional knowledge, posing higher requirements for the development of large models.
"The development of large models integrating general and specialized knowledge is becoming increasingly important. Based on the general large model, we have strengthened the knowledge of power and energy disciplines to build a large model for power and energy disciplines. It can not only serve scientific research on AI-empowered new power systems but also accompany and guide students' learning effectively," Liu Yanli explained.
Looking forward to AI assistants "blooming everywhere"
As a demonstration course at Tianjin University, the "Power System Fundamentals" course became the first stage for this "digital teacher." Soon after, this "AI assistant covering the entire process and all aspects" will be fully introduced into a more compatible course—"Smart Grid and AI."
"I mainly conduct interdisciplinary research in power and energy and artificial intelligence. I have developed this course focusing on AI empowering the new power system's digital transformation, which is at the forefront of the discipline. By transforming typical applications of AI in new power systems into student projects, students can apply AI technology to solve the core key technology research and development of smart grids," Liu Yanli told the China Science Daily. She also plans to invite international faculty to co-teach and guide the course, enabling students to learn theoretical knowledge more efficiently using AI assistants and enhance their ability to apply interdisciplinary knowledge to solve complex engineering problems.
Using AI assistants to serve interdisciplinary talent cultivation is not only Liu Yanli's original intention in creating AI assistants but also the core goal for the future.
"According to previous training models, students who want to enter interdisciplinary fields may need to minor in a second major, which takes a relatively long time," Liu Yanli said. "However, with the assistance of AI teaching, students can directly focus on research projects at the forefront of interdisciplinary engineering, greatly reducing the time required to study courses in other fields and achieving short-cycle, high-quality, and personalized interdisciplinary talent cultivation tasks." In building and guiding students in interdisciplinary research, AI teaching assistants can be incredibly helpful. "Different disciplines can collaborate through human-computer interaction to guide students in learning interdisciplinary knowledge, fostering talents with multiple skill sets," says Liu Yanli. "Moreover, interdisciplinary human-computer interaction facilitates the exploration of innovative points in cross-disciplinary research."
The creation of multidisciplinary "digital teachers" relies on the participation of more universities and educators. When asked about the promotion of "AI teaching assistants throughout the entire process," Liu Yanli expressed great enthusiasm. "After the release of our AI teaching assistant, many teachers came to me seeking advice, and I shared my research and development process with each of them."
She believes that "with the ongoing deepening of artificial intelligence technology, courses solely based on knowledge dissemination can be replaced by 'digital teacher' courses. We must embrace cutting-edge technology to transform higher education and cultivate high-level and innovative talents."
From Liu Yanli's perspective, AI teaching assistants have strong scalability, and any field can attempt to create its own "digital teacher." "I am considering turning the experience of developing AI teaching assistants into a MOOC, where my 'digital teacher' can share it with everyone! Feel free to 'copy my homework'!" she says with a smile.
