AI-powered curriculum development: What are the pros and cons research has revealed?

Contents:

• What are the risks associated with using AI in designing educational programs?
• What benefits does AI bring to learning design?
• What should be considered when implementing AI in learning design?

In the past two years, many scientific studies have been conducted on the impact of generative AI on the development of educational programs. Philippa Hardman, a former research fellow at the University of Cambridge and creator of the DOMS™ learning design system, summarized the key findings of these studies. She highlighted both the objective risks and benefits associated with integrating AI into instructional design. These findings have the potential to significantly impact the future of educational technologies and methodologies, as well as approaches to teaching and learning. It is important to consider both the positive and negative aspects to effectively integrate AI into the educational process and ensure the best outcomes for students. In this text, we will review the research findings selected by Philippa, focusing on the less obvious benefits and drawbacks associated with designing educational programs and creating learning materials using neural networks. We will not delve into well-known issues such as "hallucinations," information security risks when working with open-source neural networks, or controversial ethical and legal aspects. Instead, we focus on the practical aspects of using neural networks in education, which will allow us to better understand their potential and limitations in the context of developing effective curricula and materials.

What are the risks associated with using AI in the design of educational programs?

Modern research has identified the vulnerabilities of artificial intelligence that must be taken into account when developing educational programs based on neural networks.

Artificial intelligence provides new opportunities for learning design, increasing its effectiveness and simplifying the scaling of educational programs. However, experts emphasize that AI tools based on pre-prepared templates can significantly limit teachers in their methodological choices and creativity. This creates the risk of simplifying the educational process and reducing its individual adaptation to the needs of students.

Research conducted by scientists from Norway and the UK has shown that tools based on rigid frameworks that standardize the learning design process limit the freedom of action of methodologists. This creates a feeling of being "constrained by a framework" and complicates the adaptation of educational materials to the specific needs of the target audience. At the same time, more flexible solutions exist on the market, such as iLUKS and ChatCLD. These tools offer a structure that serves as a guide for instructional designers, but also allow for modifications, refinement, and additions to various aspects of the educational program. This approach contributes to a more effective learning design process and better alignment of materials with student expectations.

Read also:

Comparison of approaches to course development: an instructional designer using neural networks or without them. The question of which of these approaches is more effective remains relevant. Neural networks can significantly simplify the process of creating educational materials by providing tools for data analysis, automation of routine tasks, and personalization of content. On the other hand, a traditional approach without the use of technology requires deep knowledge of pedagogy and the ability to think creatively. Both methods have their advantages and disadvantages, and the choice depends on specific goals and objectives. It is important to consider how neural networks can complement traditional approaches, rather than replace them, which opens up new possibilities for instructional design.

Research has shown that artificial intelligence often creates content without taking into account methodological principles and does not facilitate the adaptation of materials to a specific context. In particular, a study conducted in China found that 78% of mathematics lesson plans developed using the GPT-4 model required significant revision before they met established educational standards and student proficiency levels. This underscores the importance of human involvement in the process of developing educational materials to ensure that they meet the requirements and needs of students.

Belgian researchers have found that AI tools are more focused on imitating instructional design than on its deep application to create high-quality content. They noted that many instructional designers often blindly accept AI recommendations without critically evaluating them. This leads to a decrease in the quality of the final product. It is important to consciously use AI in the educational process to ensure compliance with instructional design principles and achieve effective results.

Neural networks can create source text that requires careful review for compliance with instructional design principles and significant revision. Therefore, the expected time savings may be questionable or insignificant, given the need to review and refine the resulting material.

Philippa Hardman notes that the creative work of instructional designers is a contradictory situation. On the one hand, AI tools can significantly assist in idea generation. On the other hand, they tend to automate and standardize processes, which can negatively impact creativity. Overreliance on neural networks when making creative decisions can negatively impact the final result. Using AI in design requires a careful approach to preserve the individuality and uniqueness of the creative process. Research conducted in the United States has shown that instructional designers who actively use artificial intelligence in their work are significantly less likely to create original assignments for students compared to their colleagues who are less enthusiastic about neural networks. Similar conclusions were reached in a study from Belgium, which noted that an uncritical acceptance of AI suggestions leads to the creation of monotonous and banal educational materials. This underscores the importance of critical thinking and creativity in the educational process, especially when integrating new technologies. Generating educational assignments using neural networks is suitable if you are looking for a standard and templated approach to presenting educational materials. However, to achieve originality, it is important to use neural networks wisely. They can be an excellent aid in brainstorming with a team, but it is necessary to refine the resulting ideas independently. In the following, we will take a detailed look at how to effectively integrate neural networks into the educational process.

The use of neural networks allows you to create draft texts, but they require careful review for compliance with the principles of instructional design and significant revision. Thus, the expected time savings may not be so obvious, given the time required to analyze and correct the resulting material.

What benefits does AI bring to instructional design?

Modern research has identified key areas in which generative AI can be effectively applied in instructional design. Interestingly, many of these areas are also associated with certain risks. Sometimes research findings are contradictory, which may indicate the need for a more careful approach to the use of these technologies. It is important to understand how exactly to implement generative AI in order to maximize its potential in educational processes.

Artificial intelligence significantly accelerates the process of developing educational materials. Research by Korean scientists has shown that using ChatGPT can reduce lesson planning time by up to 65%. Meanwhile, Chinese researchers have developed the TreeQuestion AI platform, which enables the creation of multiple-choice test questions, significantly reducing test creation time by up to 95%. These achievements highlight the potential of AI in education and its ability to optimize learning processes.

Given these risks, it's important to note that reducing the time it takes to prepare educational content can negatively impact its quality. This occurs if AI-generated materials aren't properly reviewed and edited. If we do spend time editing them, the question arises about the actual speedup of the materials preparation process, taking these costs into account.

Artificial intelligence significantly simplifies and accelerates the process of assessing knowledge and skills. Automating test creation allows teachers to save time and conduct assessments more regularly, as shown by studies in China, while maintaining high accuracy. To generate high-quality tests, researchers recommend using Bloom's Taxonomy in prompts. It is important to remember that teachers need to verify that the questions generated by the neural network correspond to the students' level of preparation and truly measure their achievements against specific learning outcomes. The use of AI in the educational process not only optimizes time but also improves the quality of assessment, which contributes to more effective learning.

Personalized learning implies adapting the learning process to each student's level of knowledge and experience, as well as their interests and preferences in terms of pace and methods of information acquisition. Contextualization of learning is a modern trend in the field of learning and development (L&D). This approach also includes curriculum adaptation, but focuses not on individual student characteristics but on the specifics of the professional environment in which they work or will work. This method makes learning more relevant and effective, which facilitates better knowledge and skill acquisition. Education experts emphasize that artificial intelligence (AI) offers significant opportunities for adaptive learning, tailored to the individual student's level. New research supports this view. For example, customized chatbots developed based on the characteristics of student groups (the persona method) increased learning achievement in a course on information processing by 29%. The researchers attribute this positive result to adapted explanations, prompts, and assignments, which more effectively help students absorb the material compared to traditional, one-size-fits-all educational content. The use of AI in the educational process opens new horizons for improving the quality of education and increasing student motivation.

Read also:

Personalization using neural networks is considered impossible today, but in the future, Significant advances are likely to be made in this area. Over the next year, artificial intelligence and machine learning technologies will advance significantly, opening up new opportunities for personalized content. This will allow companies to more precisely tailor their offerings to user needs, improving engagement and increasing customer satisfaction. Investing in and developing neural networks will be key to achieving success in personalization.

The assertion that artificial intelligence is incapable of adapting materials to a specific context requires clarification. In fact, everything depends on the settings and the degree of human involvement in the process. To implement student-centered adaptive learning, a US research team developed the ARCHED framework. Within this framework, one AI tool is responsible for content generation, while another evaluates it against instructional design principles and learning objectives based on Bloom's taxonomy. A human controls the entire process, adjusting parameters to adapt the program to different groups of students and making methodological decisions. Thus, AI is becoming a powerful tool for creating personalized educational experiences.

Philippa Hardman notes that numerous studies confirm that AI chatbots and adaptive learning systems can effectively adapt to learners' needs in real time. These technologies are capable of analyzing students' actions and current performance, allowing them to provide personalized feedback. Based on the collected data, AI suggests relevant learning content and tasks appropriate to the level of difficulty, significantly improving the learning process and increasing its effectiveness.

While the implementation of neural networks is sometimes seen as a factor limiting creativity, Philip Hardman emphasizes that everything depends on the specialist's approach to using AI tools. When used correctly, these technologies can become a powerful creative assistant, capable of generating ideas that may be inaccessible to a single person. The creative potential of neural networks opens new horizons for professionals, allowing them to expand their horizons and find unique solutions to problems.

Research conducted in the United States shows that the use of artificial intelligence in brainstorming generates 47% more diverse ideas compared to traditional methods based solely on human team interaction. ChatGPT, in particular, demonstrates its effectiveness by offering educators a wide range of options for delivering educational content and diverse learning activities, significantly enriching the educational process. The use of AI in educational practices opens new horizons for innovative approaches and improved learning quality.

It is important to understand that the use of neural network suggestions should not be direct and immediate. This can negatively impact the depth and diversity of the educational process for students. Artificial intelligence tools provide significant assistance in the initial stages of course design. At such times, educators can use AI to generate new course ideas, develop assignment options, and create drafts of learning materials. Research shows that the use of AI in this context frees up the time and cognitive resources of educators, who can then focus on evaluating proposed ideas, selecting the most appropriate ones, refining them, and adapting them to specific learning conditions. Using AI as a tool to support the educational process can significantly improve its effectiveness and quality.

What to consider when implementing AI in instructional design

Philippa Hardman notes that current research shows that expectations for AI in the field of instructional design are often exaggerated. Nevertheless, the value of neural networks in this area cannot be denied. Artificial intelligence effectively copes with certain tasks and demonstrates significant results, while emphasizing those aspects where human contribution remains indispensable.

To effectively leverage the potential of artificial intelligence in instructional design and minimize risks to learning quality, Hardman offers the following recommendations. First and foremost, it's important to integrate AI into learning processes while maintaining control over content and teaching methods. Clear goals and objectives must be established to ensure AI serves as a supporting tool rather than a replacement for traditional methods. It's also crucial to regularly evaluate and update the algorithms used to ensure their relevance and alignment with educational standards. The implementation of AI should be carefully considered, including training for faculty and students in the use of new technologies. Thus, it is possible to leverage the strengths of AI to improve the educational process while maintaining high standards of teaching quality.

• Develop digital literacy and prompt engineering skills – properly written prompts improve the quality of generation by 58% compared to basic ones.
• Avoid AI tools that severely limit the independence of the instructional designer. Instead, choose flexible solutions that allow you to refine and adapt the results produced by the neural network.
• Entrust AI with generating drafts and versions, analyzing large volumes of data, scaling feedback, adapting content for different groups of students, and other routine tasks that should be automated. And devote the saved time to making creative, methodological, and strategic decisions, which neural networks, unlike humans, are not strong in.
• Carefully monitor the quality of generated materials and never use them "as is." It is important to ensure not only that there are no “hallucinations,” but also that the content corresponds to the methodological principles, educational goals, characteristics of the target audience, and the learning context.
• Pay attention to the diversity of educational content, avoid excessive standardization.

Check out additional materials:

• 5 skills of educational designers that are already changing due to the use of neural networks
• 5 mistakes in the implementation of AI tools in the training and development of employees
• 8 ideas on how teachers and methodologists can use the Perplexity neural network
• Strengths and weaknesses of ChatGPT in online course development
• How to create online courses using neural networks: real cases