Introduction 

Throughout my studies for the past few weeks I have been reflecting on the place that technology has in our assessments. I wanted to share about two of my most recent creations, continuations of the “Yes, and…” experimentation I wrote about last week. In order to do this, I wanted to have a scholarly conversation with AI to see how using an assessment technology may or may not enhance this product. Below is an interview I conducted between myself and ChatGPT with some creative licenses taken. I have included the prompt to frame my interaction below:

Please note: This interview was conducted with ChatGPT, an AI developed by OpenAI. Responses were generated based on prompts and guidance provided by the author.

Click to see the initial prompt that created this interview 

For my next blog post, I would like to have an interview with you, ChatGPT. In this interview, I would like to discuss my latest two creations (sandbox #9 and sandbox #10) and how assessment technologies impact these two assessment creations. I would also like to discuss more deeply with you the implications of using AI for assessment and what the benefits and drawbacks may be. I would like this to be a shared conversation with each of us taking turns to answer and ask questions. Ultimately, I would like to share about 500 words of my own thoughts in addition to what you add to the conversation. Given this information, guide the conversation to allow me to share my knowledge and current thoughts about AI as it relates to assessment. 

Our Conversation

A blog post interview with ChatGPT and Emma Thrasher

ChatGPT: Thanks for joining me today! You’ve been doing some exciting design work in your fictional Excellence in Customer Service course. Can you tell readers what you’ve been working on this week?

Emma: Absolutely. I’ve been designing two assessments focused on emotional intelligence (EQ), which is a critical skill in customer service. My learning target this week was for learners to identify their areas of strength and weakness in EQ, and to explore tools for growing in those areas. I created two assessment designs—one gamified and one more metacognitive—that explore different assessment strategies and technologies.

ChatGPT: Let’s start with the first one. What did that look like?

Emma: For the first assessment, I was given the technique of pre-assessment, the structure of small groups (5–6 learners), the platform Blooket, and a synchronous delivery method. I was excited to use a gamified experience to let learners assess themselves in a low-stakes way. By tracking their ability to answer questions in real time, they could begin to spot gaps in their understanding or ability to transfer knowledge about EQ concepts.

There was definitely a trade-off in my framework here—autonomy was somewhat reduced in favor of novelty and potential engagement. Authenticity came not from the game format itself but from the scenarios embedded in the questions.

ChatGPT: That’s a really thoughtful reflection. And what about the second assessment—how did it differ?

Emma: This one was more focused on metacognitive engagement. The technique was “assessment in a book,” with groups as the structure, Dugga as the technology, and a checklist for the response. Learners started by reading and taking an EQ assessment from a guidebook. Then, using Dugga—a digital assessment platform—they created a new version of the EQ assessment that could be used by peers who weren’t present. They also created a checklist to guide meaningful assessment design.

This approach offered a much higher degree of autonomy and a strong sense of authenticity. Learners were building something useful for others—a meaningful application of their new knowledge.

ChatGPT: I love the contrast in those two designs. As you were creating them, how did technology—especially AI—factor into your thinking?

Emma: Honestly, the most important thing for me when using AI in assessment is transparency. AI can be a powerful feedback partner—similar to asking a peer to read a draft or suggest improvements—but it should never be a hidden evaluator.

Beyond that, I think AI shines when used as a planning or administrative assistant. For example, it can find trends in assessment data or generate multiple versions of an item aligned to different learner needs. But I’m cautious—it’s easy to lean on AI for efficiency and accidentally overlook the human dynamics of learning, which are essential for something like EQ.

ChatGPT: Some educators worry that AI could depersonalize learning, while others believe it enables greater personalization. What trade-offs do you see?

Emma: I listened to a podcast episode from EdSurge (Young, 2021) this week entitled Should Robots Replace Teachers? where the host spoke with Neil Selwyn about AI and automation in education. One example stood out: facial recognition technology being used in classrooms to take attendance—essentially using surveillance to manage presence. Another example was Century AI, a system used in Europe that monitors student activity and makes recommendations for next steps in online learning.

While I can see the value in using AI to scaffold learning or personalize assessments—for example, generating leveled questions—I’m not fully comfortable with how this veers into surveillance. It’s important to distinguish between monitoring that supports learning (like using formative feedback to adjust instruction) and surveillance that exists primarily for control. I believe in using AI tools that empower both learners and teachers, not ones that replace or diminish the role of either.

ChatGPT: That’s a really helpful distinction. As a designer, what’s your current philosophy around AI-enhanced assessment?

Emma: In the assessments I designed this week, I was able to leverage these technologies to increase engagement and to offer learners a stretch task—to create something with the assistance of technology. I think that’s the sweet spot right now. Letting technology enhance the experience, without letting it drive the experience, allows the learner to remain in control.

That balance—enhancement, not replacement—is how we make sure our assessments are aligned with real-world technology use. Using frameworks like TPACK can help us ensure the technology serves the pedagogy, not the other way around.

ChatGPT: As we wrap up, what advice would you offer to others exploring AI for assessment design?

Emma: I’d say:

1. Start with your learning goals—don’t let the technology dictate your purpose.
   
2. Use AI as a creative partner, not a final authority. It can offer suggestions, not decisions.
   
3. Design for transparency—make it clear when and how AI is being used, and involve learners in the conversation.
   

Ultimately, technology should enhance learning, not eclipse the very human process of growth and reflection that assessment can foster.

ChatGPT: Thanks so much for this conversation. Your work shows how thoughtful, intentional design can create space for both innovation and integrity in assessment.

Emma’s Concluding Thoughts

Looking back at this interview, I can see AI being used as a complement here to help me frame my takeaways and to share my assessment processes in a novel and maybe more engaging way than my standard blog post. This is one of the main ways that I feel technology can enhance assessment and our designs. It can transform the mode of delivery or interaction to make a more engaging and novel experience for learners. In using AI to craft this reflection post, I encountered a few logistical challenges that are part of the trade-off when engaging with these technologies. I needed to make sure I crafted the correct prompts and provided enough information to share my intended message. I noticed that while AI shared my intended message, in the process of making my responses more like an interview, ChatGPT has added words to my responses, included a prolific use of the em dash, and has drawn some conclusions from my provided text that I am not sure I would necessarily connect back to my initial response. Technology poses interesting challenges when it comes to assessment or the learning process in general, and it is important to remain transparent and to continue to dialogue with others (not just with AI) about how this will impact our future both inside and outside of learning experiences.

References:

Young, J. R. (Executive Producer). (2021, October 5). Should robots replace teachers? [Audio podcast]. EdSurge. https://www.edsurge.com/news/2021-10-05-should-robots-replace-teachers