This is the fourth review of the 2015 short story nominees for the Nebula Awards. Spanish translation below by Daniela Toulemonde.
“Today I am Paul” by Martin L. Shoemaker is the third Nebula nominated short story I’ve read this year where the main character is an AI and the story is told via first-person (ha!) narration by that AI. Who knows whether the popularity of this theme was mere chance or a reflection of the growing concern about the rise of the machines, but this AI story, like the other two, features a friendly, helpful intelligence.
In “Today I am Paul” we enter a near-future world where an embodied android provides medical support and care for an elderly woman who is losing her memory. The AI can change its appearance, depending on the degree of information downloaded about the person it’s trying to emulate. This is a comfort to the patient, Mildred, who is drifting in and out of the past and present, seeking out various loved ones to talk to. In the course of a day, the AI emulates Mildred’s son Paul, her deceased husband, her daughter in-law, and her granddaughter, but the number of people who could be emulated via the “emulation net” is limitless as long as data is available.
We are informed that this is a new feature for androids, as is this AI’s empathy subnet, which directs the AI to avoid upsetting Mildred and positively finding ways to comfort her. The two programming nets do not always calibrate. While the AI does a very convincing job of emulating her argumentative son, its empathy net warns the AI that Mildred’s anxiety is increasing, which requires the AI to resolve the competing directives. In this conflicted space, self awareness is born, and the android develops an understanding of its programming, analysis and actions that are separate from the roles it plays. Shoemaker does a good job in creating a character that is no character and any character, but what is specifically outstanding about the story is that it convincingly depicts how the AI’s consciousness seems to inevitably emerge from programming tension. Continue reading