👾 Create Educational Games with AI
Claude Artifacts allow you to generate interactive games, educational simulations, and more simply with a prompt.
In our upcoming AI x Education webinar "It Could but Probably Can't: Avoiding Pitfalls of Current AI Model Usage in K12 Education" on August 2nd, we will be featuring Mike Hardy, an award-winning public school teacher, principal, superintendent, and current AI researcher at Stanford University. The webinar will focus on identifying specific educational tasks where GPT-style AI models might falter and the potential unintended consequences of their deployment. Participants will engage with the latest AI research and explore real-world case studies to better understand the practical applications and limitations of AI in public education. The session aims to equip educators and administrators with the knowledge to critically assess AI technologies, preparing them to make informed decisions about incorporating these tools in their curriculums as AI continues to evolve. Register now to secure your spot:
Here is an overview of today’s newsletter:
Examples of Interactive Artifacts You Can Create with Claude
Student Perspective Opinion Piece on AI Applications
Exploring Large Language Models from a Student's Perspective
Animated AI Teaching Assistants in College Campuses
🚀 Practical AI Usage and Policies
Claude AI recently introduced a free feature called Artifacts, which enables users to create standalone content such as interactive games, diagrams, websites, and more using simple prompts. It goes beyond generating code but also allows you to view and interact with what you generate. Now you can truly bring your vision to life with just your words! Claude Artifacts allow educators to create engaging resources for students to learn from and students to develop fascinating projects based on what they've learned. In this section, we'll explore some of the fascinating things you can create with Claude Artifacts.
To get started with Claude Artifacts, simply click the flask icon on the bottom left corner of the Claude prompt box to enable the Artifacts feature.
In the prompt box, you can ask Claude to create interactive simulations, games, diagrams, flowcharts, and more. Claude will generate your artifact in a dedicated panel alongside the main chat, allowing you to interact with your artifact and even publish it to share with others. You can also edit your artifacts or allow others to remix them, enabling them to make iterations or improvements upon your artifact.
Here are some examples of Claude Artifacts we've generated:
Math - Multiplication Maze Adventure
This is a fun and educational game you can create for students to practice their multiplication skills. Players navigate through the maze by solving multiplication problems. If they answer correctly, they can move up, left, right, or down to get to the yellow end zone. Players can earn 5 points for each correct answer and lose 1 point for each incorrect answer within a time limit of 2 minutes. Students can play the game and compare scores among the class.
Try out the Multiplication Maze Adventure here. Feel free to remix the artifact to tailor the game to your class or subject. Below is the prompt used to generate the game.
Prompt: Develop a React component for a 'Multiplication Maze Adventure' game. Create a colorful grid-based maze where each cell contains a multiplication problem. Students navigate through the maze by solving the problems correctly. Include a character that moves through the maze based on student input. Add a timer to make it more exciting. Use Tailwind classes for styling and consider adding sound effects for correct and incorrect answers. I want there to be a place where the student can type in their answer and if they get it correct, they can move one space up, down, left, or right. Otherwise if they get it wrong, you give them another problem to solve. The goal is to get from the green box to the yellow box of this maze. Each time they get an answer correct, they get 5 points. If they get it wrong minus 1 point. The goal of this game is to get to the yellow box.Physics - Double Pendulum Simulation
You can create online interactive simulations using Claude Artifacts, allowing students to adjust different factors and observe their effects in real-time. In this double pendulum simulation, students can adjust the initial angle of each pendulum by dragging sliders. They can also view a real-time angle graph displaying the changing angles of both pendulums as well as energy visualization graphs that show potential, kinetic, and total energy. These features help students understand how energy changes as the pendulum angles are adjusted. The interactive nature of simulations can promote hands-on learning and exploration of complex physics concepts.
Try out the Interactive Double Pendulum Simulation here. This simulation was generated by remixing an existing simulation by claudeartifacts.org.
English - Spelling Word Story Builder
Engage students with new vocabulary words or have them practice their spelling words using this interactive activity. In Teacher Mode, you can input a list of spelling words with definitions and examples. In Student Mode, students see a word list, a word bank, and a section to write a story using the words from the list. As students use each word from the word list, the word gets highlighted in green. This Spelling Word Story Builder demonstrates the vast potential of interactive activities that educators can create with Claude Artifacts.
Try out the Spelling Word Story Builder here. Below is the prompt used to generate the artifact.
Prompt: Develop a React component for a 'Spelling Word Story Builder' game. Create an interface where students see a list of spelling words they need to use in a story. Provide a text area for story writing, with real-time highlighting of correctly spelled and used words from the list. Include a word bank with definitions and example sentences. Add an illustration area where students can draw scenes from their stories or choose from pre-made images. Implement a teacher mode for managing word lists and reviewing student stories. Use a colorful, notebook-style design for the interface.If you are struggling to write a prompt to create your artifact, you can have Claude write the prompt for you! For instance, you can give it the role of a skilled prompt engineer and ask it to generate a prompt based on your ideas of what you want to create. You can also remix existing artifacts to customize them to your class. Feel free to share your Claude Artifact creations in the comment section below!
📣 Student Voices and Use Cases
Yash Yardi, a high school student from the Illinois Math and Science Academy, presents the following opinion piece exploring innovative AI use cases beyond conventional applications. His perspective aims to inspire and encourage students and educators to envision the diverse ways AI can be harnessed in the future.
How AI is Changing the World Around Us From a Student’s Perspective
By: Yash Yardi
You may be well aware of AI’s dominance in computer science, robotics, digital art/photography, and smart assistants. For years, AI has been growing exponentially in these fields, however, this segment dives into some of the more niche subjects–from creative to inventive–where AI may be underestimated.
Archaeology
Starting with an industry where the human-AI connection is kicking off its roots, machine learning is being utilized in archaeology through satellite imagery to better predict where valuable dig sites and buried structures are. This paper published on Nature.com demonstrates how this exact example is used in the Mesopotamian floodplain environment and had approximately 80% accuracy. From my perspective, archaeology is partially what writes the history books of tomorrow; though this field usually can be overlooked as complex or even dangerous for many, AI is guiding archaeologists to new findings and contributing to more insightful research and textbooks for future students.
Oceanography
Discovering what’s below the deep waters is not typically associated with AI. In conjunction with satellite imagery, AI in the ocean incorporates acoustic surveying and biological sampling for discovery purposes. Additionally, monitoring ocean acidity, pollution, and temperatures can be substantially stronger through AI. The full details of its applications are demonstrated in this review.
Art Restoration
Testing the limits of machine learning algorithms, art restoration is aimed to replace the grueling efforts of human hands replicating every stroke on an aged canvas. What revolutionizes this field is the sheer size of data behind the models, allowing the damage to be assessed and repaired quickly and in alignment with the style of the artwork. This Medium article also illustrates the approaches Oxia Palus and other startups take in the rebuilding of lost artworks. Although there is a level of experience needed to be a conserver, many can recognize the benefits of escaping tedious and precise tasks with AI to focus on more creative endeavors. However, understandably, this view is controversial as some say there is an art to restoration as well, so utilizing machine power too much may kill an art form.
Through these three examples of AI in unique studies, I hope to inspire people to pursue innovation no matter what the subject is. As a reader, you have the power to conquer any study you put your mind to!
📝 Latest Research in AI + Education
Pennsylvania State University
The Future of Learning: Large Language Models through the Lens of Students ↗️
Brief Summary: The use of ChatGPT in education shows a balance of efficiency and trust, with students using it for inquiries, content creation, and more. However, concerns about the "black box" nature of LLMs persist. The study highlights a trade-off between the convenience of ChatGPT and the risk of misinformation, especially for novice users.
Opinion: AI tutors can provide support outside regular hours, helping to reduce the stress on educators while simultaneously improving student engagement. However, the concerns surrounding the "black box" nature of LLMs highlight the importance of transparency in these systems. Students and educators need to understand how AI-generated responses are formed and the potential biases that arise from training data.
To address these challenges, it is vital to enhance the transparency of LLMs. Traditional techniques like "jailbreaking" and prompt engineering can only go so far in helping users understand how the LLM works - students and teachers must be informed about the decision-making processes underlying these models if they are expected to judge its output. Additionally, incorporating AI tutors should be coupled with strong supervisory roles, where educators guide students in critically assessing AI-generated content. This approach not only fosters trust in the technology but also encourages students to develop critical thinking skills and avoid over-reliance on LLMs.
Zhang, H., Xie, J., Wu, C., Cai, J., Kim, C., & Carroll, J. M. (2024). The future of learning: Large language models through the lens of students. arXiv. https://arxiv.org/abs/2407.12723University of Minnesota, Osaka University
Brief Summary: The integration of generative AI in higher education across the USA, Japan, China, and Mongolia reflects diverse national priorities and challenges. While the USA and Japan emphasize a human-centered approach to augment education, China focuses on innovation and security, and Mongolia seeks economic advancement despite funding constraints. A common challenge is addressing algorithmic biases and the digital divide, especially in the Global South.
Opinion: Implementing educational policies at the governmental level is challenging due to diverse national priorities, economic constraints, and socio-cultural differences that exist across countries and even across states and districts at the national level. To overcome these challenges, governments need to collaborate closely with AI experts, educational researchers, and educational boards responsible for school curricula. Such partnerships can help develop policies that are not only technologically apt but also inclusive of local educational establishments. Moreover, this collaboration can facilitate the creation of regulations that safeguard the ethical and equitable use of AI, protecting both students and educators.
Xie, Q., Li, M., & Enkhtur, A. (2024, July). Exploring generative AI policies in higher education: A comparative perspective from China, Japan, Mongolia, and the USA. https://doi.org/10.13140/RG.2.2.33314.03522📰 In the News
Tech Crunch
After Tesla and OpenAI, Andrej Karpathy’s startup aims to apply AI assistants to education ↗️
Key takeaways:
Andrej Karpathy's New Venture: Karpathy, former AI head at Tesla and OpenAI researcher, has launched Eureka Labs, an AI-driven education platform aimed at integrating AI teaching assistants into the learning process.
Initial Focus and Course Offering: The platform's first product, an AI course called LLM101n, aims to help students create their own AI systems. The course, which blends digital and physical cohorts, is still under development.
Uncertain Product Scope and Future Plans: Details about Eureka Labs' broader educational products, including whether they will extend to traditional educational settings like high schools, remain unclear. The platform has not yet demonstrated the efficacy of its AI teaching assistants.
Karpathy's Background and Involvement: Karpathy, who has a strong background in AI and education, has not disclosed any external funding or partnerships for Eureka Labs. His previous roles include leading AI teams at Tesla and OpenAI, and teaching deep learning at Stanford University.
Inside Higher Ed
Animated AI TAs Coming to Morehouse ↗️
Key takeaways:
Implementation of AI Avatars at Morehouse: Morehouse College is introducing 3-D, AI-powered avatars in classrooms, allowing students to ask course-related questions anytime. These avatars are designed to resemble professors and provide a more interactive learning experience.
Purpose and Functionality: The AI avatars, trained on professors' lectures and materials, aim to enhance students' access to information rather than replace human instructors. Students can interact with the avatars through a web browser, receiving responses that mimic the classroom experience.
Faculty Concerns and Responses: While some faculty members, like Ovell Hamilton, express concerns about AI potentially replacing jobs, the initiative's leaders emphasize that the technology is meant to support, not replace, educators. The technology is seen as a tool to make education more accessible and efficient.
Broader Context and Reactions: The use of AI in education is a growing trend, raising both opportunities and concerns. Faculty unions advocate for careful regulation and involvement of educators in implementing AI. At Morehouse, where there are no teaching assistants, the AI avatars are viewed as a complement to existing faculty roles.
“Chatgpt.” ChatGPT, OpenAI (GPT-4), openai.com/chatgpt. Accessed 23 Jul. 2024. And that’s a wrap for this week’s newsletter! If you enjoyed our newsletter and found it helpful, please consider sharing this free resource with your fellow educators!