A blind or low-vision user can get immediate, private scene descriptions before choosing human assistance.
Collections: Human still decides · Negative results
Executive brief
Accessibility value comes from giving the user control over when AI is sufficient and when a human is needed.
AI value · Demonstrated task scope in user study
VerifiedFourteen users described Be My AI support across personal and social contexts.
Qualitative study; no causal time-saving or safety metric.
Before
Contacts a sighted volunteer or trusted person and shares a camera view. → Describes the visual scene and answers follow-up questions.
After
Captures an image and asks Be My AI for a description or follow-up detail. → Evaluates the answer and either acts, reprompts, or escalates to a human for high-stakes or uncertain tasks.
Human boundary
The user decides whether to trust, reprompt, or escalate; humans retain action decisions.
Why it matters
A blind or low-vision user can get immediate, private scene descriptions before choosing human assistance.
Before
How the work ran before the change.
Blind or low-vision user
Contacts a sighted volunteer or trusted person and shares a camera view.
ControlUser explains intent and decides whether to act.
Human assistant
Describes the visual scene and answers follow-up questions.
ControlUser judgment.
What changed
A blind or low-vision user can get immediate, private scene descriptions before choosing human assistance.
Decision rightHuman moves from creator to judge
After
How the same work runs now.
User
Captures an image and asks Be My AI for a description or follow-up detail.
User
Evaluates the answer and either acts, reprompts, or escalates to a human for high-stakes or uncertain tasks.
Exception path
Users compensate for hallucinations or misunderstood context and intent through spatial memory, auditory cues, personal judgment, reprompting, orientation and mobility skills, or human assistance.
Decision authority
The user decides whether to trust, reprompt, or escalate; humans retain action decisions.
| # | Actor | Action | Control |
|---|---|---|---|
| 01 | Blind or low-vision user | Contacts a sighted volunteer or trusted person and shares a camera view. | User explains intent and decides whether to act. |
| 02 | Human assistant | Describes the visual scene and answers follow-up questions. | User judgment. |
| # | Actor | Action | Control |
|---|---|---|---|
| 01 | User | Captures an image and asks Be My AI for a description or follow-up detail. | User controls framing, prompting, and use. |
| 02 | User | Evaluates the answer and either acts, reprompts, or escalates to a human for high-stakes or uncertain tasks. | Human escalation remains available. |
Work that left the path
Human role before
Sighted volunteers supplied first-line visual interpretation.
Human role after
Users independently frame questions and judge outputs; human volunteers handle uncertainty and consequential tasks.
AI role
Decision mode: advisory perception. Describes images and answers visual questions; it does not safely navigate or make high-stakes decisions.
Visual interpretation commonly required sighted assistance.Fourteen users described Be My AI support across personal and social contexts.
CHI 2025 qualitative study. · 14 visually impaired Be My AI users plus analysis of image descriptions.
Qualitative study; no causal time-saving or safety metric.
Human assistants can clarify intent and context interactively.The study identified hallucinations, misinterpretations of social contexts, styles, and human identities, and failures to grasp or act on user intent.
CHI 2025 study. · 14 interviews plus image descriptions from participants and social media.
Small purposive sample; limitations may change with model updates.
Anti-pattern
Positioning probabilistic image description as safe navigation or identity verification.
Questions
Portability conditions
Reputation risk
medium