🧠 Dyslexia & AI-Assisted Learning Research

At xelsed.ai, we research how artificial intelligence can make learning more accessible for people with dyslexia and other learning differences.

The Challenge

Traditional coding environments present significant barriers for learners with dyslexia:

• Dense text - Code editors display walls of monospaced text
• Syntax sensitivity - A single misplaced character breaks everything
• Error messages - Cryptic technical jargon that's hard to parse
• Documentation - Long, text-heavy reference materials
• Case sensitivity - "myVariable" vs "myvariable" causes errors

"10-20% of the population has some form of dyslexia. That's potentially hundreds of millions of people who face extra barriers when learning to code."

Our Approach

P5 Editor AI addresses these challenges through AI-powered assistance:

Natural Language Input

Instead of writing code from scratch, users describe what they want in plain English. The AI handles the syntax-sensitive translation to working code.

• No need to remember exact function names
• No syntax errors from typos
• Focus on creative intent, not technical details

Voice Commands

Voice input removes the text-processing barrier entirely. Users can speak their ideas naturally, bypassing the challenges of reading and typing code.

Visual Feedback

p5.js is inherently visual - every change produces immediate graphical output. This provides:

• Instant validation - See if your idea worked immediately
• Spatial reasoning - Many dyslexic individuals excel at visual-spatial thinking
• Engagement - Creative visual output is more motivating than text output

Conversational Learning

The AI explains code as it writes, creating a dialogue-based learning experience:

• Ask "why" at any point
• Request simpler explanations
• Learn concepts through conversation, not reading

Design Principles

P5 Editor AI follows dyslexia-friendly design principles:

Typography

• Sans-serif fonts throughout
• Adequate line spacing (1.5+)
• High contrast text
• Avoiding justified text alignment

Color & Contrast

• Dark mode default (easier on eyes)
• Color-coded syntax highlighting
• Distinct visual regions
• Avoiding pure white backgrounds

Cognitive Load

• Clean, uncluttered interface
• One primary action at a time
• Progressive disclosure of features
• Clear visual hierarchy

Research Areas

xelsed.ai is actively researching:

AI Tutoring Systems

How can AI adapt its teaching style to individual learning needs? We're exploring personalized explanation styles, pacing, and reinforcement patterns.

Multimodal Input

Combining voice, gesture, and visual input to create coding environments that work with diverse cognitive styles rather than against them.

Error Recovery

How AI can interpret intent from imperfect input, gracefully handling the types of errors that are common for dyslexic users.

Assessment Methods

Developing evaluation approaches that measure understanding and creativity rather than syntax accuracy.

Why Creative Coding?

We chose p5.js and creative coding as our research platform because:

• Visual output - Plays to dyslexic strengths in spatial reasoning
• Immediate feedback - No waiting for compilation
• Creative freedom - No "wrong" answers, just different expressions
• Low floor, high ceiling - Easy to start, unlimited growth potential
• Community - Welcoming, art-focused community

Get Involved

We're always looking for collaborators, researchers, and feedback:

• Visit xelsed.ai - Learn more about our research
• Share feedback - Help us improve
• Contact us - Discuss research collaboration

References

• International Dyslexia Association - dyslexiaida.org
• British Dyslexia Association - bdadyslexia.org.uk
• p5.js Foundation - p5js.org