Experiment #2 with a New AI Writing System (Article 2/2)

Experiment #2 with a New AI Writing System (Article 2/2)

This article represents my honest perspective based on first-hand experience testing a GenAI tool, not a paid or sponsored promotion.


I recently got to test drive an exciting new AI writing tool that is simply truly next-level, and unlike anything I've experienced before, such as ChatGPT, Bard, Claude, or other GenAI systems. Although I cannot name the tool due to obvious (NDA, sort of) reasons, the interface and capabilities of this tool are truly groundbreaking. It was developed by a leading AI research institute, that has previously developed many computer vision, and visualization tools, among many others, towards humanity's intellectual progress. This tool is not available to the public yet.

I decided to give it a spin using a couple of my own published academic papers to see what it could come up with. The posts below were generated entirely by this cutting-edge AI tool summarizing key points from my research.

While it did an amazing job extracting and explaining core concepts from my papers, the unique perspectives and opinions expressed are the AI's own. I only made minor edits for clarity and flow.

I'm sharing these AI-written articles as an experiment to showcase this innovative new technology. The future is here! Let me know your thoughts on these machine-generated takes on my work in the comments.

Original scholarly work published in open access here - https://scholarworks.umass.edu/education_working_papers/6/

In the second experiment (first one here), the research article has been supplied to the "tool", with the ability to select individual sections of the article. One interesting option in this experiment is that you can supply the content of a section(s) as a prompt to the GenAI tool - which is unavailable in the first experiment. In one way, this allowed for setting the boundaries for the tool to work within the context of the research article. Hence, reducing the amount of hallucinations the tool may generate.


Prompt Literacy: A Pivotal Educational Skill in the Age of AI

Introduction (Supplied from original work. Again, impressive output, and ELI5 material!)

Before Generative AI came into being, one needs to know at least one programming language (such as Java, Python, etc) to communicate with the computer. But now, with the advent of Generative AI, a significant scientific achievement, is transforming education by creating personalized learning experiences and interactive content, and this interaction is facilitated through a common language called prompts, making the skill of crafting these prompts - termed as 'prompt literacy' - an emerging necessity in the 21st century.

Prompt literacy is a pivotal educational skill, becoming a necessity as we increasingly interact with generative AI systems such as ChatGPT, BingAI, etc. Prompt literacy deals with the art of crafting, interpreting, and analyzing AI prompts, is transforming the way we approach education, making it a powerful tool for the digital age.

Understanding and mastering prompt literacy not only enhances our interaction with AI but also fosters critical thinking, problem-solving, and creativity, skills that are highly valued in the digital age. Prompt Crafting Techniques In our exploration of prompt literacy, we utilized AI-powered tools, specifically ChatGPT-3.5, which was freely available to the general public. All generative AI tools are "prompt-dependent". Prompts can be used to control the content, style, or format of the generated output. For instance, a prompt for ChatGPT might be "Write a poem about a cat," which would instruct the AI to generate a poem about a cat, leaving room for creative freedom in terms of length, rhyme scheme, or meter. While this sample prompt does generate a poem about a cat, it is too general. This is where the CAST model comes into play. CAST stands for Criteria, Audience, Specifications, and Testing. It provides a framework for crafting effective prompts, ensuring that they are specific, relevant, and tailored to the task at hand.

For demonstration purposes, let's take the example of a "create a lesson plan" prompt. Initially, it might sound like "create a lesson plan for teaching introduction to magnetism for 5th graders". When applied the CAST model, it becomes:

Criteria: The lesson plan should introduce the concept of magnetism, its basic principles, and real-world applications.

Audience: The lesson plan is intended for 5th-grade students, so the language and examples used should be age-appropriate.

Specifications: The lesson plan should include learning objectives, teaching methods, activities, and assessment methods.

Testing: The effectiveness of the lesson plan can be tested by assessing the students' understanding of the concept of magnetism after the lesson.

Such precise and to-the-point prompts not only generate more accurate lesson plans, but also produce less hallucinations. We encourage educators and students to experiment with the CAST model further. Implications for STEAM Education The key takeaway from our work is the critical role of prompt literacy in the age of AI. As AI becomes more prevalent in classrooms, it's not just about generating text, but also creating images, videos, presentations, and other multimedia content. However, the effectiveness of these AI tools is fundamentally dependent on the quality of the prompts provided. Therefore, educators must adapt and prioritize this skill to ensure successful AI integration.

One of the most interesting aspects of our work was discovering how small changes in the wording of a prompt can produce significantly different results. This highlights the importance of being precise and ready to revise prompts to maximize the information obtained from AI systems.

AI can be used in classroom learning in innovative ways, both in formal and informal settings. For instance, educators can use AI to enhance teaching methods and create deeper learning experiences. They can also promote collaborative learning activities where students work together to create prompts for AI tools, thereby fostering teamwork, problem-solving, and critical thinking abilities.

Moreover, AI can serve as a powerful tool for underserved communities where there is less reach of education and literacy initiatives. By leveraging AI, we can broaden educational outreach and promote social good. For example, AI can be used to create personalized learning experiences, making education more accessible and engaging for students in remote areas.

Potential use cases include: In underserved communities, AI can be a game-changer in literacy initiatives. For instance, AI-powered applications can be used to deliver personalized learning content to students in remote areas. These applications can adapt to the learning pace and style of each student, making education more engaging and effective.

Moreover, AI can be used to translate educational content into various languages, making it accessible to non-native English speakers. It can also be programmed to incorporate culturally relevant examples in teaching materials. For instance, in teaching mathematics, an AI tool can use examples of local markets or traditional games familiar to the students, making abstract concepts more relatable and easier to grasp.

AI can also facilitate remote learning, a necessity in areas with limited educational infrastructure. Through AI-powered platforms, students can attend virtual classes, interact with teachers, and collaborate with peers, all from the comfort of their homes. This can significantly increase the reach of education, bringing learning opportunities to students who would otherwise be left behind.

AI can be used to track and analyze student performance via stealth assessments, providing valuable insights to educators. This data can be used to identify learning gaps and tailor teaching strategies accordingly, ensuring that no student is left behind.

For teacher education and professional development, AI can be used to highlight and address biases in educational content. For example, an AI tool can analyze textbooks and identify instances of gender or racial bias, helping educators create a more equitable learning environment.

For cultural and history educators, AI can be used to create virtual reality experiences that allow students to explore different cultures and historical periods. This not only enhances their understanding of the subject matter but also fosters empathy and respect for cultural diversity. (Unexpected, but truly a great resource for history teachers!)


As we navigate the 21st century's AI revolution, the importance of prompt literacy cannot be overstated. It is rapidly establishing itself as a key skill, holding significant promise for transforming education, particularly within STEM and non-STEM fields. The ability to create effective prompts will determine the success of AI integration, making it imperative for educators to adapt, personalize, and prioritize this skill.

The future of education lies in the hands of educators who are willing to embrace this change and adapt their teaching methods to harness the power of AI. As we continue to explore the potential of AI in education, we must also focus on equipping our educators with the necessary skills and knowledge to navigate this new landscape.

In the future, we hope to see more research and development in the area of prompt literacy, focusing on creating more effective and efficient prompts. We also look forward to seeing how AI can be further integrated into the classroom to enhance teaching and learning experiences.

Although not covered in this blog the potential pitfalls and challenges, we need to establish robust data privacy and security measures to protect student information at all times, and we will leave it to the experts in the cyber security and data privacy to lead the efforts (I prompted it to expand to other areas of interest; originally not in the scholar work). To all educators and future leaders of the world, we wish you good luck as you embark on this exciting journey of integrating AI into your classrooms. Remember, the key to success lies in your ability to adapt and evolve. So, keep learning, keep experimenting, and keep pushing the boundaries of what is possible.


I'm thoroughly impressed by this new AI research tool (experiment 2) after testing it on my papers. It accurately extracted key information and themes, transforming them into coherent, easy-to-understand blog articles. This represents a real advance in context-aware AI writing aids, especially for learners who are ESL.

Until next time.


Did you find this article valuable?

Support Sanskriti Blog by becoming a sponsor. Any amount is appreciated!