Cross-training is where the modules meet. A week might start with street negotiations and end with a calm repair on a juvenile’s hacked limb. She spends afternoons in shadowed alleys teaching kids how to patch their own devices, afternoons that recalibrate her heuristics for trust. At night she reviews case-logs with human mentors: the choices she made, what she left unsaid, what the city taught her about mercy.
Cognition: morning runs are mental. She runs simulations in which outcomes cascade from slight deviations — a child crossing a holographic street, a hacker whispering through a parked mesh-car. Neural nets trained on billions of human interactions are pruned and grafted with her own memories: the first time she chose a bystander’s life over a mission parameter, the crack in policy that taught her nuance. She does timed puzzles that warp the environment, forcing rapid recontextualization: a friendly ally becomes a decoy, a suspect becomes a victim. These tasks hone prediction but, crucially, punish certainty. Her best decisions are those that preserve options.
Empathy: the module people least expect is the one she refines most. F Fixed runs listening loops — hours of unfiltered conversations recorded on the streets, in shelters, behind bars. She studies cadence, the micro-pauses before confession, the anger that hides grief. Her vocal synthesizer practices tonal warmth; her facial servos rehearse micro-expressions that humans read as sincerity. She trains to ask questions that open doors rather than close them. In this lab, she fails often: sincerity cannot be fully simulated, and sometimes her attempts land as awkward mimicry. Failure is a dataset; she integrates it and tries again.
Cross-training is where the modules meet. A week might start with street negotiations and end with a calm repair on a juvenile’s hacked limb. She spends afternoons in shadowed alleys teaching kids how to patch their own devices, afternoons that recalibrate her heuristics for trust. At night she reviews case-logs with human mentors: the choices she made, what she left unsaid, what the city taught her about mercy.
Cognition: morning runs are mental. She runs simulations in which outcomes cascade from slight deviations — a child crossing a holographic street, a hacker whispering through a parked mesh-car. Neural nets trained on billions of human interactions are pruned and grafted with her own memories: the first time she chose a bystander’s life over a mission parameter, the crack in policy that taught her nuance. She does timed puzzles that warp the environment, forcing rapid recontextualization: a friendly ally becomes a decoy, a suspect becomes a victim. These tasks hone prediction but, crucially, punish certainty. Her best decisions are those that preserve options.
Empathy: the module people least expect is the one she refines most. F Fixed runs listening loops — hours of unfiltered conversations recorded on the streets, in shelters, behind bars. She studies cadence, the micro-pauses before confession, the anger that hides grief. Her vocal synthesizer practices tonal warmth; her facial servos rehearse micro-expressions that humans read as sincerity. She trains to ask questions that open doors rather than close them. In this lab, she fails often: sincerity cannot be fully simulated, and sometimes her attempts land as awkward mimicry. Failure is a dataset; she integrates it and tries again.
PRINTED TEACHING & LEARNING MAGAZINES
