JSXGraph AIR Buddy
+++ JSXGraph AIR Buddy is currently in the prototype stage and will be available for download at a later date. +++
Talk to Your Construction
JSXGraph AIR Buddy is a conversational companion in the JSXGraph AIR family.
While JSXGraph AIR for VS Code supports authors who work directly with code, AIR Buddy removes the code from view entirely: you create and control interactive mathematical constructions by simply describing them — typed or spoken (via your OS), in natural language.
Behind the scenes, an AI assistant “translates” your words into JSXGraph code and renders the result instantly as a live, fully interactive construction in the browser. You stay in the mathematical conversation; AIR Buddy takes care of the programming.
AIR Buddy is aimed at learners, teachers, and authors who want to explore, demonstrate, and experiment with dynamic mathematics — without writing a single line of JavaScript. Even JSXGraph board properties can be controlled using natural language.
Examples
Creating a construction by text or voice
You describe, AIR Buddy constructs:
🗣️ “Draw a triangle with a circumscribed circle.”
A triangle with its circumcircle appears — draggable, dynamic, and correctly constrained: move a vertex, and the circumcircle follows.
🗣️ “Now add the three perpendicular bisectors.”
The bisectors are added to the existing construction, meeting — as they must — in the circumcenter.
Interacting by text/voice
The conversation doesn’t end once the construction is on screen. You can keep steering it:
🗣️ “Color the circumcenter red and label it M.”
🗣️ “What happens if the triangle becomes obtuse? Move C so we can see it.”
Every instruction modifies the live construction in place. Exploring a conjecture becomes a dialogue: ask, look, drag, ask again.
Bidirectional control: AI – Code – Speech/Text
AIR Buddy is not a one-way text-to-picture generator. All three layers stay connected in both directions.
Speech/Text <=> JSXGraph Code <=> Live Construction
- Language to Construction: Spoken or typed instructions are translated by the AI into JSXGraph code and rendered immediately.
- Construction to Language: Changes you make by hand — dragging a point, moving a slider — feed back into the conversation. You can ask AIR Buddy about the current state: “What are the coordinates of C right now?”, “Is the triangle still acute?”
- Code in the middle: The JSXGraph code is the shared, precise representation both sides agree on. It is always there, always up to date — whether you look at it or not.
The code layer is hidden by default
By default, AIR Buddy shows only the conversation and the construction. The generated JSXGraph code — the intermediate layer between language and graphics — is deliberately kept out of sight, so that nothing distracts from the mathematics.
But it is never out of reach: one click reveals the live code panel. This makes AIR Buddy work at every level of expertise:
- Learners interact purely through language and the construction itself.
- Curious users can peek at the code and discover how their words became JSXGraph elements — a gentle, motivating entry point into programming dynamic mathematics.
- Authors and developers can open the code layer permanently, edit it directly, and watch construction and conversation stay in sync.
Didactical Perspectives
JSXGraph AIR Buddy provides a foundation for further research in mathematics education, enabling the investigation of AI-supported natural language interaction, accessibility, multilingualism, and mathematical learning.
Mathematical Construction through Natural Language
- How effectively can learners construct mathematical objects using natural language instructions?
- How does the use of everyday language compare to formal mathematical language when creating mathematical representations?
- What types of ambiguities arise when learners describe mathematical objects in natural language, and how can they be resolved?
- How does speech input differ from text input in supporting mathematical construction and exploration?
- How does natural language input support the construction of both geometric objects and function graphs?
Multilingualism and Language Diversity
- How well can learners construct mathematical objects using their preferred language?
- To what extent does multilingual input reduce language barriers in mathematics education?
- Can learners successfully interact with mathematical software using a language that is not their first language?
- How does AI-supported multilingual interaction benefit multilingual learners and students with migration backgrounds?
- How do language proficiency and mathematical proficiency interact when using natural language interfaces?
Accessibility and Inclusive Mathematics Education
- How does language-based interaction improve accessibility in mathematics learning for students with motor impairments?
- How does natural language input compare with mouse-, touch-, or gesture-based interaction in terms of accessibility and learning outcomes?
- Which learner groups benefit most from language-based mathematical interaction?
- Can language-based interaction reduce barriers for students with disabilities or special educational needs?
Mathematical Learning and Reasoning
- How does constructing mathematical objects through language influence students’ mathematical reasoning?
- Does verbalizing mathematical constructions improve conceptual understanding?
- How does AI-assisted mathematical construction affect students’ understanding of mathematical concepts and representations?
- To what extent does language-based interaction foster mathematical communication, argumentation, and problem solving?
- How does switching between natural language and symbolic mathematical representations influence learning?
Human–AI Interaction in Mathematics
- How do students interact with AI when refining, correcting, or extending mathematical constructions?
- What role does AI feedback play in supporting mathematical learning processes?
- How much guidance should an AI system provide to maximize learning while preserving productive struggle?
- How do learners perceive trust, transparency, and agency when using AI-assisted mathematical tools?