Skip to content
AI for Everyone
Menu
All articles

AI BOOST Level 3: When students take up the baton

Collaboration

7 min readBy Juliusz Tomeczek

Kolaż slajdów prezentacji: fragment tekstu polecenia (promptu) dla AI o Krakowie, ilustracje postaci w stylu grafiki AI, plansza z listą odcinków podcastu/serii („Etapy realizowania projektu”, tematy odcinków) oraz logo Koła Naukowego Geopolityki i Geografii Ekonomicznej. Na środku duży pikselowy napis „AI BOOST LEVEL 3”. W prawym dolnym rogu logo AI for Everyone z adresem www.aiforeveryone.blog.

January 16, 2025 will go down in my memory as one of those special days that confirm that sharing knowledge makes profound sense. At the recent AI BOOST meeting at the Cracow University of Economics, I saw something that filled me with real pride - students who, just weeks after starting their AI adventure, are able to use these tools in a truly creative and professional way.

Projects that delight

Analysis of media sentiment toward geopolitical conflicts

The first group surprised us with an ambitious analytical approach. Using ChatGPT to generate code in Python, they created a tool to analyze the sentiment of online articles about global conflicts. This shows how AI can support serious social and geopolitical research.

A slide titled Wojna domowa w Jemenie, meaning the civil war in Yemen, with a bar chart of the number of articles by conflict and sentiment label. Three groups of bars: Houthi rebels 48 and 48, the Saudi-led coalition 30 and 67, humanitarian aid 8, 81 and 9.

Result of an app exploring the sentiment of articles on armed conflict “Civil War in Yemen”.

Create a book on your own terms

Another team demonstrated the true synergy of AI tools - from ChatGPT to DALL-E to speech synthesis.

A screenshot of a PowerPoint window showing a slide headed Etapy realizowania projektu, meaning the stages of carrying out the project. On the left an abstract graphic in turquoise and red, on the right seven numbered steps in Polish, from writing the prompt for the chat, through describing the characters and the plot, to generating the text, the images and the voice-over.

Project implementation plan.

Translation of text from the image:

Project Implementation Stages

  1. Creating a prompt for the chat to assume the role of an adventure story author.
  2. Initial character descriptions based on the people participating in the project.
  3. Defining the event plan (story outline)
  4. Text generation
  5. Using a website that helps paraphrase text generated by chat
  6. Generating images for the story
  7. Generating an AI narrator

They created a full story about their group, enhanced with illustrations and an audio track. This is a great example of how AI can support creative multimedia projects.

Two AI-generated illustrations side by side. On the left, a colourful comic-style scene with a group of young people in front of a classical building, one of them holding an unfolded map. On the right, a sepia scene: young people in an old workshop hall among machines, with a typewriter and newspapers in the foreground.

The result of working with the DALL-E model.

Science-based cookbook

This group particularly surprised me. Not only did they use ChatGPT, Claude AI, Copilot and NotebookLM to create content, but they also reached out to SCISPACE, Elicit and Litmaps and Typeset to find a scientific basis for their recipes.

A light slide in two parts. The first says that every team member prepared ten recipes using AI tools, after which ChatGPT merged and standardised them into a single cookbook. The second lists the tools used: Copilot, Claude, ChatGPT, Notebook LM, Litmaps, Elicit and Typeset.

Scope of work.

Translation of text from the image:

Preparation:

Each team member used AI tools to develop 10 recipes.

Then, through ChatGPT, we combined and standardized all files into one cookbook.

Tools used:

  • Copilot;
  • Claude;
  • ChatGPT;
  • Notebook LM;
  • Litmaps;
  • Elicit;
  • Typeset

This shows a mature approach to combining creativity with academic rigor.

A slide headed Przykład śniadania, meaning a breakfast example. On the left, the prompts given to ChatGPT and Claude asking for a cookbook of ten recipes in Polish and English with calorie counts. On the right, a finished recipe for oatmeal with fruit and nuts: ingredients, six preparation steps and nutritional values, all in both languages.

Example of one of the recipes, tasty :)

Translation of text from the image:

Breakfast Example

ChatGPT: Take on the role of a student creating a final project for a course on using artificial intelligence. Generate a cookbook containing 10 breakfast recipes in Polish and English, don’t forget to include the caloric values of the ingredients.

Claude: Based on these files, generate a cookbook in two languages - Polish on the left and English on the right, provide the caloric value of all ingredients, there must be exactly 10 recipes. Don’t forget about the step-by-step preparation method.

  1. Oatmeal with Fruits and Nuts

Ingredients:

  • Oatmeal (50g)
  • 2% Milk (200ml)
  • Banana (1 medium)
  • Blueberries (50g)
  • Walnuts (10g)
  • Honey (1 teaspoon)

Instructions:

  1. Bring milk to a boil in a pot.
  2. Add oatmeal and cook on low heat for 5 minutes, stirring frequently.
  3. Transfer to a bowl.
  4. Slice the banana.
  5. Top with sliced banana, blueberries, and chopped walnuts.
  6. Drizzle with honey.

Nutritional values:

  • Calories: 495 kcal
  • Protein: 15g
  • Fat: 16g
  • Carbohydrates: 78g

Plan of the podcast

A project that went beyond our course.

A slide headed Proces tworzenia, meaning the creation process, with five steps joined by downward arrows. Coming up with the podcast name, logo and description in ChatGPT, finding articles in Elicit and SciSpace, an episode outline in NotebookLM, the script and ideas for further episodes in ChatGPT, and finally a soundtrack in Suno AI.

Creation process.

Translation of text from the image:

Creation Process

  1. Creating the name, logo and podcast description
  • Tool: Chat GPT
  1. Finding articles for episode script
  • Tool: Elicit, SciSpace
  1. Creating episode outline
  • Tool: NotebookLM
  1. Creating episode script and proposals for future episodes
  • Tool: Chat GPT
  1. Creating podcast soundtrack
  • Tool: Suno AI

The team not only prepared a comprehensive plan for the 6 episodes, but also learned on their own how to use Sunto AI for sound production. This shows that we gave them not only the tools, but also the ability to develop on their own.

A slide headed Propozycje kolejnych odcinków, meaning proposals for further episodes, with five cards. Each has a generated cover and a title: episode two on trade routes, three on climate and rivalry between powers, four on a technological cold war, five on an African renaissance, six on the geopolitics of raw materials.

Plan for future episodes of the podcast.

Translation of text from the image:

Proposals for upcoming episodes

  • Ep. 2.
    • “Trade Routes 2.0: From the Silk Road to Arctic Passages”
  • Ep. 3.
    • “Climate Under Scrutiny: Ecological Rivalries Between Superpowers”
  • Ep. 4.
    • “Technological Cold War: The Battle for Chips and Data”
  • Ep. 5.
    • “African Renaissance: Myth or Coming Reality?”
  • Ep. 6.
    • “Geopolitics of Resources: How Natural Resources Shape the World?”

Inclusion quiz

.

The last group created an interactive quiz using ChatGPT to help visitors discover their “spiritual neighborhood” of Krakow.

A screenshot of a chat with a model on a dark background. The prompt, in Polish, tells the model to act as a guide to the districts of Krakow, ask ten open questions one at a time, and after the last answer assign the person a matching district together with a justification, a description and a humorous summary.

Developed prompt.

Translation of text from the image:

You are an expert on Krakow and a guide to its districts. Have a conversation with me that will help me discover which district of Krakow I would be if I were a place. Ask me 10 open questions one by one, one question at a time. Each question should be about my lifestyle, how I spend my time, my approach to people, and preferences regarding place of residence. Conduct the conversation naturally – ask a question, wait for my answer, and only then ask the next one.

After my answer to the 10th question:

  • Assign me a district of Krakow that best matches my answers.
  • Provide a common summary of my answers and explain why you chose this particular district.
  • Add a brief, educational description of this district, including its character, interesting places, and the lifestyle that prevails there.
  • Create a street that could symbolize my lifestyle in this district.
  • Add a humorous description of my character as this district, referring to my answers.

This is a great example of how AI can serve social and inclusive purposes, with a touch of humor and local color.

A fragment of the model answer on a dark background, headed Humorystyczny opis Twojego charakteru jako dzielnicy, meaning a humorous description of your character as a district. The model describes the person as a calm, well organised district of Krakow where life runs slowly, and ends by asking whether such a district suits them.

Fragment of an answer provided by ChatGPT as part of the quiz.

Translation of text from the image:

Humorous description of your character as a district:

If you were a district of Krakow, you would be a calm but well-organized place where life flows leisurely and without unnecessary rush. Guests visit you willingly, but on your terms – first theater, then kebab, and finally a movie. Neighbors don’t bother you, as they’re either young and busy, or not there at all. In your world, everything is functional and comfortable, with greenery and benches always ready for an afternoon with a book.

Does such a district suit you? 😊

Technology bonus

Using the remaining time, I showed the students some more news from the world of AI:

  • The function of projects in ChatGPT
  • Creation of personalized GPT’s
  • PERPLEXITY as an alternative to Google

Final Reflections

I found the implementation of the AI BOOST project an extremely inspiring experience. Seeing how the students went from the basics to advanced applications of AI in such a short period of time, I am convinced that the future of the technology in the Polish academic environment looks bright.

Thank you also for recognizing my contribution to this project

Photo of a framed letter of thanks from the Geopolitics and Economic Geography student association at the Kraków University of Economics for running nine hours of AI BOOST training in November and December 2024 and in January 2025.

Translation of text from the image:

Krakow, 16.01.2025

LETTER OF APPRECIATION

The Scientific Club of Geopolitics and Economic Geography operating at the Department of International Relations of the Cracow University of Economics extends sincere thanks to

MR. JULIUSZ TOMECZEK

For willingness to share his experience and conducting 9-hour AI BOOST training sessions for the Scientific Club of Geopolitics and Economic Geography at the Campus of the Cracow University of Economics on 14.11.2024, 12.12.2024, 16.01.2025.

[signature] President of SC GEG

From my side, thanks are due to the entire Scientific Circle of Geopolitics and Economic Geography of UEK for the great cooperation and commitment to this project. I believe that this is just the beginning of your adventure with AI.

Group photo in a lecture room: about twenty training participants stand in a row, with the presenter in the middle holding a framed letter of thanks. A projection screen, a whiteboard and an Executive MBA banner are visible behind them.

Note!
This article was written in collaboration with Claude 3.5 Sonnet, an advanced language model from Anthropic. As an AI-conscious user, I use the latest technology to create valuable content, while maintaining full control over the form and content of the publication.

This article has also been automatically translated from Polish using DeepL. If you find any errors, please let me know in the comments.

Juliusz Tomeczek

About the author

Juliusz Tomeczek

My name is Juliusz Tomeczek. I run the AI for Everyone blog and publish articles here about artificial intelligence.

More about me