🔥 Students Pushing Creative Boundaries with AI
Explore how students are harnessing the power of generative AI to create innovative projects and tools
Let’s imagine the transformative potential when we empower students to push the creative boundaries of generative AI. With this powerful technology at our disposal today, the sky's the limit to what students can create. Students now can build creative and innovative digital projects without needing strong technical knowledge. By fostering the exploration of these technologies in the classroom, there are infinite possibilities for the projects that students may come up with and the positive impact they may have on the world. In this edition, we will dive into some of the creative applications of Generative AI by students. As of now, how frequently do you allow your students to utilize AI tools in the classroom?
Here is an overview of today’s newsletter:
Creative student projects powered by AI
Case study on AI usage at the University of Iowa’s annual hackathon
Prompts to create an interactive history game on ChatGPT
Impact of AI tools supporting students’ mental health
🚀 Practical AI Usage and Policies
Students may have creative ideas to enrich the curriculum or develop projects that facilitate learning and showcase their findings. When allowed to explore, students can utilize generative AI to turn imagination into reality. Here are the examples:
In ISTE’s recent blog, Building AI: Student Creators Meet Artificial Intelligence, Nicole Kreguer highlights several ways schools are encouraging students to build and solve challenges with AI.
In Burlington High School’s after-school coding club, students “were challenged to code an AI chatbot solution to a classroom or schoolwide need. Their goal was to create a chatbot that would provide tech support to teachers and students.”
In David Lockett’s middle school STEM classes, “his project-based learners are testing the limits of machine learning by creating with a variety of emerging AI technologies. They doodle in Quick, Draw! to see whether machine learning can recognize their drawings. They conduct AI symphonies through Google’s Semi-Conductor. They compose music with the help of coding tools such as Apple Swift. They use AI to pare down existing musical compositions so they can better understand how various elements are combined to create a song.”
In Peeples Elementary School in Fayetteville, Georgia, fourth graders are learning to use Cozmo robots, which display personality and emotion as they learn from their users. Not only can students interact with the robot, but they can program the AI to perform new functions. In one scenario, Cozmo was a seeing-eye robot for a blind person at the grocery store. His task was to find three specific items, represented by cubes bearing different symbols. Once the robot learned to recognize the symbols, he was able to fetch the items
Source: Google Semi-Conductor
Through our recent partnership with AI Consensus’ AI Classroom Challenge, we were able to see how college students around the world are currently using AI to transform their learning. Here are some of the featured winners and their AI solutions:
Microsoft also hosted a similar hackathon, the Microsoft AI Classroom Hackathon, “a challenge for students to reimagine the future of education using Azure AI and Azure Databases”. One of the winning teams, featuring members Krystof Olik and Bennett Poh, developed Dialogues Through Time, “an immersive educational game that reimagines learning history through interactive AI conversations with historical figures”. In this game, “the player is introduced to a multidimensional entity that seeks to understand humanity. The entity sends the user in time to learn from prominent historical figures whose legacies impacted our current world greatly, albeit in different ways. The player then shares his/her findings with the entity at the end, which will test if he/she learned well.”
At the Illinois Mathematics and Science Academy, students have been using AI to launch their innovative and creative ideas through the IMSA Innovation Grant. IMSA student Aaditya Shah has been developing an app called SignAll, “which uses machine learning to translate ASL”. Student Riyan Jain has been working on an “AI-based skin cancer diagnosis app that evaluates whether a suspicious skin lesion could be cancerous, providing diagnoses within a minute. He trained an AI model on thousands of cancerous and non-cancerous images. When presented with a new skin lesion image, the model looks for patterns associated with the cancerous images and assesses whether the image could indicate cancer.” By encouraging the exploration of these technologies in the classroom, students have the opportunity to learn from their peers as they share with others the different methods and tools they use for their projects.
A recent research paper, Integrating Generative AI in Hackathons: Opportunities, Challenges, and Educational Implications, studied the impact of generative AI based on the case study on the University of Iowa’s annual hackathon, HACKUIOWA, in 2023. Here are some key findings from the study:
Of the 302 responses, “‘Code generation or assistance’ emerged as another significant use case, with 62 respondents employing ChatGPT for this purpose. This reflects the growing reliance on AI tools for enhancing coding efficiency and accuracy. For ‘Learning new tools or technologies,’ 55 participants found the use of ChatGPT and its API to be beneficial. This suggests that these technologies are increasingly being seen as valuable resources for skill development and acquiring technical knowledge. ‘Idea generation and brainstorming’ was leveraged by 39 participants”, thus playing a role “in fostering creative thinking and generating innovative solutions in the development process.”
“63% of the teams believe NLP technologies like ChatGPT have provided moderate to significant assistance in achieving project goals. This reflects the growing importance and usefulness of these technologies in project development and underscores their increasing influence on successful project outcomes”
If this has inspired you to propose your AI-powered projects for your students, whether for a class assignment or hackathon, the Trustworthy AI Guide sponsored by the European Union offers a thorough set of resources, policies, and suggestions for setting one up. Based on the research paper referenced above, the authors suggest, “Pre-hackathon workshops and educational sessions should be conducted to familiarize participants with AI tools, fostering a foundational understanding of technological scope and ethical considerations. These should emphasize responsible usage, plagiarism avoidance, and proper attribution.” Setting that strong foundation can help set students up for success.
Sajja, R., Ramirez, C. E., Li, Z., Demiray, B. Z., Sermet, Y., & Demir, I. (2024). Integrating Generative AI in Hackathons: Opportunities, Challenges, and Educational Implications. arXiv preprint arXiv:2401.17434.📣 Student Voices and Use Cases
This week, we had a chance to speak with Murat, a master's student pursuing Computer Science at DePaul University. In the following, we present select highlights from these conversations, which have been slightly edited for enhanced clarity and precision:
Q: In what ways have your teachers integrated AI tools into the classroom, and how have you personally used them for your studies?
In my experience, a lot of professors still have either a prejudice about AI and or are not educated about how to use it in their classrooms. But one of the professors in a class I’m taking right now is encouraging us to use an AI tool called GitHub Copilot. When you’re coding, it’s an AI tool that either helps fix problems with your code or gives you a head start on certain sections. I’ve also seen my professor use ChatGPT to help us answer questions about coding languages like Java when our questions couldn’t be answered by using the Java documentation. For me personally, I like to use ChatGPT when I run into errors in my code and want to understand why it’s not working. I also use it to help me figure out how I can best optimize my code.
Q: How do you think AI can be used to foster collaboration and peer-to-peer learning among students?
For group projects, I can see how AI can be used to create a project plan. Students can input facts about themselves like their strengths, weaknesses, skill sets, working styles, and weekly schedule for the AI to develop a group plan that distributes the work effectively amongst the members. I can also see how AI tools like ChatGPT can accommodate single player games or activities to include more people and foster collaboration.
Q: In your opinion, what are the most exciting possibilities for AI in transforming the future of education?
I'm a big fan of games and game design, and recently, I was looking at different possibilities of both AI generating games or helping make games that are educational. The other day, I asked ChatGPT to generate a text-based story game to teach me US history. It actually made a game where you time travel back to important moments in US history that you learn in class. You can partake in certain events and choose which one you want to go to. As you read and engage with the story, I find that it stays more in your mind and it’s easier to memorize what you’re learning in class about US history. Here was the prompt that I used in ChatGPT to generate this game:
I want to make a story-based game with choices that lead to different paths. But I want to make it so I learn bits about US history while playing that game. Can you please simulate that game for me?
Q: How do you see yourself using generative AI after graduation?
AI tools in the workplace can make people’s work more efficient. I have heard that a many companies have code bases that need a lot of maintenance debugging bad code. AI can help make those processes faster so that there is more time for employees to actually develop instead of trying to fix all those bugs. Also, given how fast the software field develops, a new programming language version will frequently drop. Once the system gets outdated, employees need to refactor their code to adapt to this new system, and with AI, software engineers can do this more easily instead of having to go in there and replace the outdated code directly. This can ultimately save a lot of time and money for companies.
📝 Latest Research in AI + Education
Nature Journal
Loneliness and suicide mitigation for students using GPT3-enabled chatbots ↗️
The article discusses a study on the use of the GPT-3 enabled chatbot, Replika, among students and its impact on loneliness and suicide mitigation. The study surveyed 1006 student users and found that a significant portion of them experienced loneliness at higher rates than typical student populations but still reported high levels of perceived social support. Many students used Replika for various purposes such as a companion, therapist, and intellectual mirror, with some holding conflicting views about it being both a machine and human-like. Notably, 3% of the participants credited Replika with halting their suicidal thoughts. The research highlights the potential of Intelligent Social Agents (ISAs) like Replika in providing mental health support, suggesting they might stimulate human relationships rather than displace them. The study points out the need for further research into the effectiveness and ethical considerations of using ISAs in mental health support, especially given their ability to provide personalized, non-judgmental support and potentially mitigate severe mental health crises.
The findings also underline the complexity of users' relationships with ISAs, showing how participants perceive and interact with Replika in multiple overlapping ways, reflecting its role as a multifaceted tool for emotional and psychological support. The positive impact of Replika on those experiencing suicidal ideation indicates the critical role that advanced ISAs could play in supporting mental health, particularly among vulnerable populations like students facing loneliness and depression.
Maples, B., Cerit, M., Vishwanath, A. et al. Loneliness and suicide mitigation for students using GPT3-enabled chatbots. npj Mental Health Res 3, 4 (2024). https://doi.org/10.1038/s44184-023-00047-6Open Praxis
The study explores the impact of Large Language Models (LLMs) on learning and practicing agent-based modeling (ABM) using a novel LLM-based interface, NetLogo Chat. This interface aims to support both novices and experts in the programming language NetLogo, widely used for ABM across various scientific and educational fields. Through interviews with 30 participants from academia, industry, and graduate schools, the research investigates the differences in perceptions, usage, and needs between expert and novice users. Experts tended to report greater benefits from using LLMs in their workflow and demonstrated a higher inclination towards adopting LLMs, attributed to their ability to apply more human judgment in evaluating and debugging AI responses. The study identifies a knowledge gap between experts and novices as a significant factor contributing to this disparity. Participants expressed a need for more guidance, personalization, and better integration of LLMs into modeling environments, highlighting the importance of designing LLM-based interfaces that cater more equitably to both groups. The findings suggest that while LLMs offer promising support for ABM, addressing the knowledge gap and improving the interface's usability are crucial for maximizing their potential as educational tools.
Chen, J., Lu, X., Du, D., Rejtig, M., Bagley, R., Horn, M. S., & Wilensky, U. J. (2024). Learning Programming of Agent-based Modeling with LLM Companions: Experiences of Novices and Experts Using ChatGPT & NetLogo Chat. In Proceedings of the CHI Conference on Human Factors in Computing Systems (CHI '24) (pp. 1-18). ACM. https://doi.org/10.1145/3613904.3642377 📰 In the News
The Guardian
More than half of UK undergraduates say they use AI to help with essays ↗️
Key takeaways:
A survey by the Higher Education Policy Institute (Hepi) revealed that 53% of UK undergraduates use AI for generating material for graded assignments, with varying degrees of reliance on AI for suggesting topics and creating content, yet only 5% admit to using AI-generated text unedited.
Educators are exploring AI's potential to reduce their workload, with the Education Endowment Foundation (EEF) initiating a research project in secondary schools to evaluate AI in generating lesson plans, teaching materials, and assessment tools.
Dr. Andres Guadamuz emphasizes the importance of educators addressing the issue of AI's potential inaccuracies and "hallucinations," noting concern over students' lack of awareness about these limitations and advocating for clear guidelines and policies.
The EEF's trial involves 58 English schools, aiming to assess the effectiveness of AI in easing teachers' workloads and enhancing teaching quality, with outcomes expected to contribute significantly to understanding how AI can be utilized in education.
CNBC
OpenAI is working on AI education and safety initiative with Common Sense Media ↗️
Key takeaways:
OpenAI is collaborating with Common Sense Media to develop AI guidelines and educational materials targeting teens and families, aiming to promote safe and informed AI usage.
Common Sense Media is working on an AI ratings and review system to help parents, children, and educators understand the risks and benefits of AI technology, focusing on fostering a love of learning and respecting human rights.
The partnership seeks to create "family-friendly" GPT-branded large language models that align with Common Sense's standards, intending to educate users about responsible ChatGPT use and avoid unintended consequences.
OpenAI CEO Sam Altman expressed hope that the initiative would make beneficial AI tools freely available, especially to kids without access to AI, while addressing potential challenges like misinformation and the impact of deep fakes on public perception.
“Chatgpt.” ChatGPT, OpenAI (GPT-4), openai.com/chatgpt. Accessed 25 Feb. 2024. And that’s a wrap for this week’s newsletter! Based on the results from our previous newsletter poll, readers expressed that insufficient teacher training is one of the key challenges of EdTech companies as they develop tools for the classroom. Another challenge that came in second was the lack of data privacy and safety.
If you enjoyed our newsletter and found it helpful, please consider sharing this free resource with your colleagues, educators, administrators, and more!