Should Initial teacher training include how to use AI

The Promise and Peril of AI in Education

As new technologies emerge, there’s often a temptation to view them as revolutionary forces that will transform the nature of education. However, the fundamental principles of teaching and learning have remained remarkably consistent over time, even as tools like writing, the printing press, and the internet have reshaped how knowledge is transmitted and accessed.

The introduction of large language models (LLMs) such as GPT-4 is the latest development to spark excitement and debate within the education community. While these AI systems hold immense potential, there are also valid concerns about their use, particularly when it comes to high-stakes student assessment. As an experienced IT professional and a lecturer in initial teacher training, I believe it’s crucial that we approach the integration of AI in education with a thoughtful, nuanced perspective.

Preparing New Teachers for the AI Era

One of my primary responsibilities is training the next generation of teachers at the University of Roehampton. As I consider how to best prepare these individuals for the realities of the modern classroom, the role of AI cannot be ignored.

The current standards for qualified teacher status and the Initial Teacher Training (ITT) core content framework make no explicit mention of technology, despite the fact that new teachers will undoubtedly encounter AI-powered tools and applications in their day-to-day work. This oversight needs to be addressed, and the next iteration of these guidelines must include clear guidance on how to use AI effectively and responsibly.

My colleagues and I have been particularly concerned about the potential for academic misconduct, with students submitting work generated by AI rather than their own original efforts. We’ve implemented policies to address this issue, emphasizing that unattributed use of generative AI would be treated as plagiarism. However, the challenge extends beyond simply preventing misuse – we must also help our students develop the skills to use these technologies wisely and ethically.

Leveraging AI to Enhance Teaching and Learning

While the use of AI in high-stakes assessment is fraught with risks and limitations, there are several promising applications of this technology that warrant careful exploration. One area where LLMs can offer significant value is in providing low-stakes, formative feedback to students.

By leveraging the natural language processing capabilities of models like GPT-4, educators can offer immediate, targeted feedback on specific aspects of student work, such as grammar, spelling, and punctuation. LLMs can also highlight areas for improvement in the structure, clarity, or coherence of a piece of writing, and offer prompts to encourage deeper reflection or analysis.

Crucially, however, such AI-generated feedback should be treated as a supplement to, rather than a replacement for, human feedback. Educators should carefully review and contextualize the input provided by AI systems, and create opportunities for students to discuss, unpack, and apply that feedback in dialogue with their instructor. Used judiciously, AI-powered feedback could encourage students to take greater ownership over their learning, engaging in self-assessment and revision of their work.

AI may also have a role to play in supporting adaptive learning and personalized support for students. By analyzing patterns in student performance data, LLMs can identify specific skills or concepts that a student is struggling with, and recommend targeted resources or activities to address those individual learning needs. However, the implementation of such systems must be approached with great care, ensuring that they are designed to enhance, rather than replace, the role of human educators in diagnosing learning needs, building relationships, and guiding student growth.

The Limitations of AI in Assessment

Despite the potential benefits of AI in education, the notion of using LLMs to grade student work is fundamentally problematic. While these models may generate feedback that appears thoughtful and thorough, they are ultimately producing probabilistic outputs based on patterns in their training data, without the capacity for genuine reasoning or understanding.

Arguing that “it’s better if you use a better prompt” is a flawed premise because it fails to address the core limitations of LLMs in the context of evaluating student work. An LLM may generate feedback that seems appropriate on the surface, but it lacks the deeper insights, critical thinking, and subjective judgment that a human educator brings to the grading process.

Furthermore, the use of AI in high-stakes assessment raises concerns about fairness, accountability, and transparency. While human graders may exhibit biases, they can be trained to recognize and mitigate them, whereas an LLM’s biases are inherent to its training data and architecture, and may be more difficult to identify and address. The appearance of objectivity in AI-generated grades may mask underlying disparities and hinder efforts to ensure equitable evaluation.

Towards a Balanced Approach to AI in Education

Rather than seeking to automate the assessment process, we should focus on leveraging AI to support and enhance human educators’ abilities, while preserving the essential role of human judgment and expertise in evaluating student learning. This may involve using LLMs for tasks like generating formative feedback or personalizing learning experiences, but with careful oversight and integration into a broader, human-centered approach to education.

As we navigate the integration of AI in the classroom, it’s crucial that initial teacher training programs equip new educators with the knowledge and skills to use these technologies responsibly and effectively. This includes understanding the fundamental limitations of LLMs, recognizing the potential for bias and inconsistency, and developing strategies for mitigating the risks associated with AI-powered assessment.

Ultimately, the introduction of AI in education is not a simple binary choice – it’s a complex, nuanced challenge that requires careful consideration, ongoing research, and a commitment to preserving the core values and practices that have sustained effective teaching and learning throughout history. By approaching this integration with a balanced, informed perspective, we can unlock the potential of AI to enhance the educational experience, while safeguarding the essential role of the human teacher.

Conclusion

The emergence of AI-powered tools like large language models has undoubtedly introduced new opportunities and challenges within the education sector. As an experienced IT professional and a lecturer in initial teacher training, I believe it’s essential that we approach this integration with a thoughtful, balanced perspective.

While there are certainly promising applications of AI in education, such as providing formative feedback or supporting personalized learning, the use of these technologies in high-stakes assessment remains fundamentally problematic. The inherent limitations of LLMs, including their inability to truly understand and evaluate student work, as well as the potential for bias and inconsistency, make them unsuitable for replacing human judgment in the grading process.

As we prepare the next generation of teachers, it’s crucial that initial teacher training programs equip them with the knowledge and skills to use AI responsibly and effectively. This includes understanding the limitations of these technologies, recognizing the potential risks, and developing strategies for integrating AI-powered tools into a broader, human-centered approach to education.

By taking a measured, informed approach to the integration of AI in the classroom, we can unlock the potential of these technologies to enhance the educational experience, while safeguarding the essential role of the human teacher. It’s a delicate balance, but one that is essential to ensuring that the future of education remains rooted in the timeless principles of effective teaching and learning.

Practical Tips for Using AI in Initial Teacher Training

As you integrate AI into your initial teacher training program, consider the following practical tips:

1. Emphasize Critical Evaluation: Ensure your students understand the fundamental limitations of LLMs, including their inability to truly comprehend and evaluate student work. Encourage them to approach AI-generated outputs with a critical eye, verifying information and assessing the reliability of the system’s responses.

2. Promote Responsible Use: Develop clear policies and guidelines for the appropriate use of AI in academic work, emphasizing the importance of attribution and the need to maintain academic integrity. Provide training on how to use these technologies effectively and ethically.

3. Explore Formative Applications: Investigate how LLMs can be used to provide low-stakes, formative feedback to students, with the understanding that this should supplement, rather than replace, human-provided guidance and assessment.

4. Address Bias and Equity: Be mindful of the potential for AI systems to exhibit biases and exacerbate existing inequities in education. Develop strategies for identifying and mitigating these issues, ensuring that the use of AI does not disadvantage certain groups of students.

5. Empower Students: Equip your students with the knowledge and skills to understand how AI systems work, how to prompt them effectively, and how to critically evaluate the outputs. Encourage them to develop the qualities that make them uniquely human, such as creativity, empathy, and critical thinking.

By adopting a thoughtful, balanced approach to the integration of AI in initial teacher training, you can help prepare the next generation of educators to navigate this evolving landscape and harness the potential of these technologies while safeguarding the core principles of effective teaching and learning.