WHAT WE HAVE DONE THIS YEAR 23/24 III: AI IN THE EXAM PART 2: what do we do with AI?

In this post on 7 June and this video last Friday, I started this story by sharing some of my thoughts on the role of AI in my subject exam.
As I told you then, in addition to asking them to use AI to “kick-start” the exam and to tell me how they had used it, in the last part of the exam (the “metacognitive” part), I asked them to reflect on

“Your utilization of AI in developing this exam and your assessment of how your humanity complemented this process”.

I confess that, now that I think about it, the formulation of the human “complementing” was not the most accurate, but well… the thing is that I would like to share with you some of the answers that stand out -I confess that I have only taken answers from people who did a minimally coherent job and that I think reflect that use in the result of their exams- and that I have organised around 5 concrete ideas:

1. AI’s Role in Enhancing Learning and Efficiency AI has significantly enhanced the planning, creation of didactic materials, and overall learning experience for students by ensuring efficiency, accuracy, and personalized learning activities. However, the human role in interpreting AI proposals and making decisions based on experience and judgment is crucial.

“The use of AI has been crucial during the exam as it has helped me planning the activities, creating the didactic materials and making the activities more fun and dynamic in order to enhance students’ learning. Also, AI ensures efficiency and accuracy and tailors the activities to meet individual learning needs.
However, it’s important to acknowledge that I have played an important role in the creation of this project. My experience, judgment, and knowledge has allowed me to interpret the AI proposal, consider the broader context of student learning, and make important decisions of my students well-being..”

2. Limited AI Usage and Human Effort: While AI was used for specific tasks like creating rubrics and translating texts, the bulk of the work remained human-driven. The AI was mainly utilized for initial suggestions, and the human touch was essential in finalizing and personalizing the work, reflecting a general mistrust in AI’s ability to replace human input fully.
As reflected in the following paragraphs

“Apart from the part where we had to use the AI, I have only used the AI for the creation of the rubrics using PopAI and DeepL as translator. I have to admit that I didn’t modify much of the work that PopAI provided me with and I simply selected the rubrics that suited me best and discarded the ones that I found useless. Other than that, the rest of the work is entirely human.”

“In my case I have not used AI much to take this exam. I have only used it to base the creation of my evaluation rubrics for the activities I have created and to translate and summarize some texts. In any case, I have never been able to literally say what the AI says because I don’t trust it very much that it is true, plus, I always like to give my personal touch to everything I do and I never like to copy things.”

“According to the last aspect I’m proud to say that the three activities were my own idea and I didn’t use any AI to ask for them, and what I really used the AI “Chatgpt” for was to ask things once the activity was thought, as for example how to get to “universe” in Little Alchemy 2 in a short way, and what aspects should teachers evaluate on the activities for the rubric, but I also used myself, like my thoughts, and everything for asking my classmates what should we do, we resolved questions with each other instead of looking for them so I think that putting things in common is more human than searching things that could be wrong.”

3. Balancing AI Assistance and Critical Thinking: The use of AI, particularly ChatGPT, was balanced with critical thinking and personal effort. AI served as a consultative tool to assist in learning without hindering personal critical thought. The emphasis was on using AI for clarifying, rewriting, and providing initial ideas, with the final output being heavily influenced by human creativity and critical analysis.

“Another form of balance is relevant in relation to my use of AI in developing this exam. ChatGPT has been my preferred form of AI throughout this process, and I have worked to use the tool in a manner that is assisting my learning, not taking away my own thinking. That has looked like clarifying questions, asking AI to rewrite sentences or pose an initial question to, and then change and build my response from there. I believe that my humanity complemented this process in that I never want my use of AI to be hindering me from learning. I do not want this technology to be a crutch that takes away from my own critical thought. I have used AI to serve in a consultative role, assisting but not replacing my responsibilities as a student and as a person. As I navigate beyond this semester, I am determined to continue embracing new ideas, critically analyzing information, and actively exploring, ensuring that my life-long learning remains dynamic and responsive to the world around me.”

“Yeah, so, in summary, artificial intelligence got my project off the ground and technology in general was essential to fill the skills I didn’t have. But my human skills and creativity enabled me to turn it into something personal, original, and focused.”

“Lastly, regarding using ChatGPT for educational purposes, even with clear prompts, ChatGPT tends to provide general rather than educational insights. Education involves emotions and cognitive aspects that AI tools may not fully grasp. It’s crucial to consult ChatGPT for ideas, but relying solely on its information without critical reconstruction can lead to misunderstandings.”

4. AI as a Supplement to Human Decision-Making and Personalization: AI generated ideas and suggestions, particularly for student profiles and activities. However, the final decisions were made based on human judgment and experience. Personalizing educational content was a key aspect where human insight was indispensable, ensuring that activities were engaging and relevant to students.

“I’ve used ChatGPT to ask for and receive suggestions. That’s to say I asked ChatGPT if I was right, if I could do this thing in a more appealing way, if I could do this other thing in a more attractive/engaging way. We all know that AI is wonderful and it can do everything you want it to do. However, I think that ChatGPT is not going to be able to do some things and being a teacher is one of those. My humanity completed this process by thinking in the things that I would have loved to do when I was a student. I think that being a teacher implies having like another sense. And you have to bear in mind always that you once were there, you once were a student and I would have loved to create a StopMotion when I was learning the European countries, for example, or to use Augmented Reality to see the different pieces of art all along history. So, I think it’s really important to never forget that you once were a student and to do what you think your students are going to enjoy.”

“As far as AI is concerned, I have used it especially to add ideas to the characteristics and abilities of my imaginary students, something that has been really useful for me as it has contribute to the amplification of the profile of my learners, allowing me to see some aspects that didn’t come into my mind at the moment…but also to develop my original idea for the activity. At every moment, and as we have had the opportunity to see in the initial video, I have tried to bring my personal and human touch to this process by changing or concretizing certain aspects proposed by the AI.”

5. Ethical Use of AI and Human Emotional Support: The ethical use of AI is paramount, as it can easily overstep into areas that should be driven by human morals and creativity. AI lacks the capacity for empathy and emotional support, which are essential in education. The human role in maintaining moral standards, providing emotional support, and ensuring the learning process remains effective and genuine is highlighted as crucial.

“The AI has been able to help me a lot in this exam to organize myself to take it in the best possible way and to give me some ideas that have been very useful to me. But it is obvious that artificial intelligence does not have the capacity to express empathy or comprehension, and emotional support that only humans can offer effectively.”

“The most important one from my point of view was the correct utilization of AI. We used it several times during this course, including during this exam. AI can be a very dangerous tool, as it can be used to create entire projects, essays, works, etc. Because of that, it is essential to have our morals in mind. If we use AI in the wrong way, the process of learning will be completely lost, that’s why during the realization of this exam I tried to use the AI for only inspiration purposes, as I wanted to test how my abilities were and how far my imagination for creating this project could go. If we use it in the right way, AI can be one of our best friends when creating and developing everything we can imagine.”

“And finally, the last question, that is, what kind of AI I have used to develop this project and also how my humanity has complemented this process. Well, I have used Schemely and Magic School to develop this project. Schemely to ask for different activities or different ideas of activities that I could develop. And Magic School to develop and create my students’ evaluation rubric. I think that my humanity has complemented this process because maybe the AI can give you many answers, but there are no answers if there are no questions. And for example, in Schemely I took into account the necessities of my students, the age they were and the contents that I wanted to teach them, I was the one deciding if I wanted to use that activity or not. And in magic classroom, I was the one who put the criteria that I wanted, the marks that I wanted, I was the one designing everything. But AI just helps us.”

“Also, this course has helped me to understand that artificial intelligence is very useful but it is only a tool or a help. Our humanity is crucial to our jobs and especially in primary school teaching. In this project I have done a lot of different questions to Chat GPT. For example, about the interest of my children but then I have to make a lot of changes to best fit with my beliefs or tastes.”

Many of their testimonies make me optimistic, and others remind me of clichés that we have read in 40 of the thousands of reports on AI that have appeared worldwide over the 1.5 years since ChatGPT appeared, but they continue to make me think.

For now, we close the review of what we have done this year with this flavour of mouth that should open the appetizer to put more ideas on the table next year.

WHAT WE HAVE DONE THIS YEAR 23/24 II: AI in the exam part 1 or reluctance to go back and forth

Last year, we introduced AI into the subject’s activities. In fact, as I told you in a video last year, the implementation strategy had been good, and I thought it was the right way to go.

This year I felt I should continue along the same path, so I tried to continue with this idea. In all the assignments of the course, AI was introduced in one way or another as an “advisor”, to brainstorm initial ideas or to develop resources with a more or less careful prompt refinement process.

However, in an attempt to be coherent, the exam had to be aware of AI and appropriate it in a more complex process. That is why I introduced Resources and ICT in the exam.

How did I introduce it? Well, I did it in two moments:

This was the first part of the exam, which is worth half of the exam I was telling you:

Your mission is to develop a teaching project for an imaginary group of primary school students. The project should incorporate technology to enhance learning. Your task is to design an assignment where students will actively engage with at least three different technologies, with at least two of them being chosen from the following options: Comic/Images, Augmented Reality, Stop-Motion, Robots/Programming. The aim is for students to actively create using these technologies rather than passively consume content. You have the flexibility to create one comprehensive activity or three –or more– connected activities, depending on your preference.

  1. Then, first, please describe the Initial Details:
  • Age of Students: Specify the age group of your imaginary students.
  • Curriculum Goals: Define the educational objectives aligned with the Spanish (or your original country’s) core curriculum that students will achieve through this project.
  • Student Profile: Describe the characteristics and abilities of your imaginary students.
  1. AI Inquiry: With these initial details, consult your preferred AI tool what to do (please record this interaction in a video and show it in your video exam). And then, before continuing with part 3, justify -in general- if you will do changes of the AI proposal or not.
  2. Project Execution: And now, for every part of the project:
  • Clearly outline the steps students will follow to complete the assignment.
  • Show in the video how to do the task and how the final artefact (product) looks.
  • Explain the added value of incorporating digital tools into the project.
  • Detail how you will assess the task and provide a rubric for grading the assignment.
  • Assess the chosen apps/tools (all of them) based on their privacy, data protection, and rights policies following the Rubric for eLearning Tool Evaluation? Address the concerns identified.
  • Utilize the ACAD Framework to articulate the learning design of your project, either by parts or in its entirety. You can use the physical toolkit or the virtual one.

As an “exam,” my students sent me this exercise in video format and also sent me the transcript, including references, detailed laws, and so on.

And what happened?

Well, I have the feeling that AI has harmed the performance of a sector of my students that is precisely the one most in need of other types of help… Let me see if I can explain myself clearly:

There are some brilliant, VERY good, VERY well-done exams for which the AI helped them to improve their perspective a lot… They are those students who are very bright and hard-working and who go all out in their exams… the AI has been an interesting resource that has enriched their work, although I confess that less than I expected.

But that’s not the majority of the student body. What happened to the others?

Well, what I should have foreseen would happen: First of all, the students “translated” the assignment, not in language, they simplifyied it (yes, that’s the theory I always keep in mind, in theory)… but instead of doing it themselves, most of them asked ChatGPT (or their favourite AI, there are several) to “translate” my exam for them (What do I have to do if my teacher asks me for…”, and what did the AI do? The predictable thing is that it turned the exam wording into its simplest and most superficial formulation, skipping all the details and nuances that make a final exam a final test of competencies…

And what did the students do? They took that simplification and, without changing almost anything, put it back into the AI and asked it to give them the answer…

The result? Yes, a disaster… very nice videos, but with “clay legs” underneath… no contextualisation, no objectives or competences to develop, generic students, generic apps in generic classes, kids who do nothing and apps that only decorate the classroom… almost always standardised… but with little foundation.

In short! little humanity… a lot to think about…

I also asked them to reflect a bit on the role of AI in their exam… but that’s another story, and this post is already too long, so in two weeks, I’ll tell you about it…