Expectations change how the same experience feels.
More quotes“Was Iris responsible?”
After the simulated crash, participants who heard the soothing voice assign the system less blame.
That is fascinating and slightly alarming.
A voice can help people understand an unfamiliar system by announcing what the vehicle detected and confirming an action. A name makes the interface easier to address and remember.
Trust and justified reliance are different results. If a human-sounding system receives forgiveness after failure, anthropomorphism may blur accountability exactly when the user needs a clear account of what happened. A friendly voice can make limited software seem considerate, aware and competent beyond its actual capability.
That gap matters well beyond vehicles. A finance assistant with a name may receive disclosures people would withhold from a plain form. A health interface that says “I understand” may imply comprehension beyond its capability. The warmer the interaction feels, the more important it is to state what is processed, retained and passed to a person. Comfort must never silently enlarge consent.
The design also contains a gender issue. Giving service systems female names and soothing female voices can repeat assumptions about who assists and reassures. “Customer preference” deserves examination rather than a shrug.
The standard should be calibration. Does the human cue help the person predict what the system can do, notice uncertainty and intervene at the right time? Or does it merely raise comfort scores? Test behaviour after errors as well as stated trust before them. State limitations in ordinary language. Keep responsibility with the organisation that designed and deployed the system.
After the crash, the customer asks what the sensors detected and who controlled the next action. Support opens the event log.