Access AI is an accessibility-focused AI project designed to help blind and low-vision users understand, read, identify, and interact with the world around them.
Our goal is to make useful visual information available naturally, quickly, and with as little friction as possible.
Access AI is being developed to explore how modern multimodal AI can make everyday visual information more accessible to people who are blind or have low vision.
Ask questions about objects, surroundings, products, clothing, signs, and other visible information.
Use AI to help read signs, labels, packaging, menus, documents, and other printed information.
Move toward natural voice-first interaction instead of requiring users to constantly operate a screen.
The current Access AI prototype demonstrates the foundation. Future development will focus on making the experience faster, more conversational, more accessible, and increasingly hands-free.
Quickly describe relevant surroundings without overwhelming the user with unnecessary information.
Identify everyday items and answer contextual questions about what the user is viewing.
Help interpret written material such as labels, packaging, signage, menus, and documents.
Explore AI-assisted search for specific items in a user's environment.
Maintain context so users can ask follow-up questions naturally rather than restarting every interaction.
Reduce dependence on touchscreens through short spoken commands, audio feedback, and wearable controls.
The next major phase of Access AI is exploring a hands-free visual assistant built around developer-friendly smart glasses.
Rather than continuously transmitting unnecessary video, Access AI is being designed around intelligent visual modes. Normal use can remain idle until requested, while reading, object-finding, scene description, and other specialized modes can temporarily analyze the visual information needed for the task.
Access AI is being developed by Anthony Pope, an independent solo developer who is visually impaired himself.
Anthony attended the Governor Morehead School for the Blind in Raleigh, North Carolina and has spent much of his life around people with different levels of blindness and visual impairment.
As his own vision continues to decline, the goal behind Access AI is deeply practical: create the kind of visual-assistance technology he would personally want available in everyday life.
That lived experience also creates an opportunity to develop and eventually test Access AI with other blind and low-vision users in real-world situations.
Access AI is currently seeking conversations with smart-glasses manufacturers, accessibility organizations, developers, researchers, and technology partners interested in helping evaluate hands-free AI-powered visual assistance.
We are particularly interested in developer hardware, evaluation units, official SDK access, technical collaboration, pilot programs, and opportunities to test Access AI in real-world accessibility environments.