Marketplace All-in-One AI's role in improving accessibility
Nov 28, 2025
Taylor Arndt, Chief Operations Officer at Techopolis Online Solutions, has been blind since birth and is a passionate advocate for inclusive AI. She shares her personal journey of overcoming accessibility challenges, from using a scanner and screen reader in school to teaching herself coding at 14. Arndt discusses how AI has the potential to improve accessibility but warns that without inclusive design, it may also perpetuate inequities. She emphasizes the necessity for AI to be trained on properly accessible examples to better serve all users.
AI Snips
Chapters
Transcript
Episode notes
Self-Taught Accessibility Began With A Scanner
- Taylor Arndt taught herself to scan printed homework and convert it to text so she could complete school assignments independently.
- That early workaround led her to learn coding at 14 to solve accessibility problems for others.
Partner Coding Beats Blind Trust In AI
- Taylor Arndt describes coding with AI as partner coding where she instructs and reviews the output instead of fully relying on AI.
- She warns many AI-generated websites lack basic accessibility like labeled buttons and form fields.
Training Data Determines Accessibility Outcomes
- AI trained on inaccessible examples perpetuates inaccessible outputs, making AI a potential unequalizer.
- Including curated accessible coding examples is essential for models to learn correct accessibility practices.