Скачать с ютуб Transforming Science Education with Generative AI Frameworks, Biases, and Future Prospects в хорошем качестве

Transforming Science Education with Generative AI Frameworks, Biases, and Future Prospects

Artificial Intelligence (AI) has made remarkable strides in recent years, finding applications in various fields, including science research and industry. Its integration into STEM education has gained attention more recently, particularly with the advent of generative AI (GAI) tools. However, technologies are not always developed with education as a primary focus. As the use of GAI among educators and students grows, it is essential to connect these technologies to learning theory, pedagogy, and classroom practice. In this seminar, I will stress the knowledge and skills teachers need to integrate generative AI in their classrooms critically and effectively. Additionally, I will discuss GAI through the lens of Technological Pedagogical Content Knowledge (TPACK) and a taxonomy for evaluating Open Educational Resources (OER). I will emphasize two key points in the first part of the seminar. First, teachers’ TPACK is crucial for using GAI in an accurate and responsible manner. Second, prompt engineering — crafting instructions for AI tools — requires competencies beyond TPACK’s technological dimension. In this regard, I will stress the importance of integrating AI-specific knowledge, such as awareness of bias, discrimination, and hallucinations. These points will be illustrated with three examples of using ChatGPT in the context of chemistry education. In the second part of the seminar, I will discuss evaluation methods. I will stress that integrating AI tools in education offers potential pedagogical advantages, such as aiding lesson planning, fostering personalized learning, and enhancing student autonomy. Yet, concerns about bias and discrimination associated with these technologies are increasing. Currently, there are no standardized evaluation criteria to assess these tools’ educational value and reliability within teaching and learning contexts. This gap, can be addressed by adapting an existing evaluation taxonomy to better align with the distinct features of GAI. This adapted approach introduces a six-dimensional evaluation framework, covering descriptive, pedagogical, representational, communication, scientific content, and ethical and transparency dimensions. This framework is applied to analyze the educational potential and ethical considerations of 30 AI tools, providing a critical mapping of both the opportunities and risks associated with AI-powered technologies in educational settings. I will summarize the seminar by discussing the knowledge teachers need to apply GAI effectively, highlighting the need to develop further theoretical frameworks for teachers’ knowledge in the age of GAI.