39. Phosphene Font
- Status: Research/speculative
- Interface point: Retina or visual cortex
- Carrier: Depends on interface
- Transducer: Electrode, optogenetic target, or other neural interface
- Encoding style: Learned symbols
- Addressing method: Stable percept map
- Expected percept: Simple symbols, arrows, warnings, or letters
- Main limitations: Training, fatigue, percept variability, limited vocabulary
Summary
A phosphene font treats artificial vision as a learned symbolic interface. It
does not try to create a normal picture. Instead, it defines a limited alphabet
of percepts the user can recognize reliably.
Chain Map
scene or message
-> symbol selection
-> stimulation pattern
-> phosphene symbol
-> learned interpretation
Example Vocabulary
| Symbol Class |
Use |
| Arrow |
Direction of obstacle, path, or motion. |
| Pulse |
Warning or attention cue. |
| Line |
Edge, step, doorway, or text baseline. |
| Cluster |
Object location. |
| Sweep |
Movement or scanning direction. |
| Simple glyph |
Letter, number, or mode indicator. |
Engineering Workarounds
- Make symbols sparse and high-contrast.
- Use repeated training games to build fluency.
- Pair each symbol with audio or tactile confirmation at first.
- Use the same symbol consistently across tasks.
- Avoid pretending it is natural sight.
Open Questions
- What is the smallest useful symbol set?
- Which percepts are stable across users?
- How much training creates automatic interpretation?
- Could a font be personalized like a prosthetic fit?