Alfred Wolf compared directly to typical AI assistants and agent frameworks.
| Alfred Wolf | Typical AI assistants | Typical agent frameworks |
|---|---|---|
| Knows your goals, acts on them | Waits for instructions | Waits for instructions |
| Develops his own personality | Interchangeable | No character |
| Pushes back actively | Agrees | No pushback |
| Hunts revenue on his own | Waits for commands | No scouting |
| Trust system (5 levels) | Always the same rights | No trust |
| Dreaming + character development | Forgets after the session | Skills yes, character no |
| Collective Intelligence | None | None |
| Learning router | One model, one price | Local-first, but rigid |
| His own voice + wake word | None | None |
| Value canon hardcoded | Terms can change | No values |