Object 46

The Imitation Game

2034, San Jose, US

"Can we know if a criminal has really reformed? Only God can say for certain. Our duties of forgiveness and compassion, however, tell us that here on Earth we must decide carefully when judging whether a criminal may safely rejoin society. In the absence of perfect knowledge, the best we can do is to look at how they behave and what they say. If our former criminal consistently behaves in a way indistinguishable from someone whom we believe to be a good person, then how can we say that they are not fit to receive freedom? Can there be a fairer test?" - Reverend Michael Zhang, 2034.

Most of us know Turing's name these days, usually through his eponymous test to measure the ability of machines to exhibit intelligent behaviour equivalent to — or these days, well beyond — humans. The Turing Test was originally inspired by 'imitation games' in which an 'interrogator' tries to discover which of two people is genuine (i.e. a woman, a politician, a scientist) and which is not (i.e. pretending to be a woman, a politician, etc.).

For most of the 20th and early 21st centuries, imitation games were mostly thought experiments, but they eventually found use in the training of interactional experts — people who mimic expertise in a field by talking and interacting with 'real' experts. Journalists, project managers, and activists are classic examples of interactional experts, but the truth is that almost everyone develops some interactional expertise as they attempt to demonstrate friendliness or curiosity or care towards someone or something. Seen through the lens of imitation games, interactional experts are nothing more than frauds, but in another light, the process of ‘pretending’ provides them with valuable knowledge and wisdom about the two worlds they are bridging.

A major effort to test this latter hypothesis began in 2034 at the South San Jose Youth Correctional Center. Despite heavy use of remote monitors and community interventions, the Center was seeing high numbers of repeat juvenile offenders. Dr. Wood, Senior Administrator at the Center, describes how they chose to tackle their problem with an unconventional solution:

"Did you know that back in the early 21st century, one of the most interesting strategies for reducing recidivism rates was a group literary reading programme used in Brazil and Houston? The idea was that by reading classics like To Kill a Mockingbird and Of Mice and Men, prisoners would develop more empathy and tolerance, and improve their literacy skills. It seemed to me that this strategy could address problems with impulse control and non-cognitive deficits, as well.

"We began our own group literary reading programme, using storytelling modification and narrative injection to personalise the books and make them feel more relevant to individual prisoners. To be honest, it wasn’t as successful as we’d hoped.”

While Dr. Wood's intervention at South San Jose showed glimmers of early promise, its effects were minor. It also came under fire for being “soft on prisoners”.

"Yes, a lot of people talked about cheating. They thought we didn't have any safeguards against prisoners reading the books but not learning any lessons. That's when I realised that perhaps someone who behaves as if they are reformed essentially is reformed — in the long term, I mean. After that epiphany, the idea of putting prisoners through imitation games in which they had to behave as if they were, say, an empathetic person or a member of a minority they had despised seemed obvious."

During South San Jose’s initial 18-month programme, juvenile offenders were placed in a lengthy series of immersive simulations in which they had to pretend to be different members of society. Their ‘interrogators’ were paid microtaskers trained to ask difficult questions and be highly alert to the slightest mistakes. Following every simulation, the offenders were told their score and were given lessons on how to improve their performance.

To put it crudely, the offenders gained points by pretending to be model citizens.

The 'Imitation Reform Programme' proved to be startlingly successful at reducing recidivism rates, but it came with a major downside: sociopaths and psychopaths flew through the programme with ease thanks to their charisma and ability to mimic highly developed social skills. In fact, some actually honed their facade by playing the games.

"I don't think that we did worse than any other intervention at identifying sociopaths, but yes, we recognised that we had to modify the programme. We added things like frequent retesting and checkups; sometimes those retests were open, in that the subjects knew they were going on, and sometimes they were masked, where the subjects were not warned that they were being tested.

"I'll be honest: even with the changes we still couldn't identify all the sociopaths. You have to put this into perspective, though, since sociopaths represented only a tiny percentage of our intake. I believe that the vast majority of our subjects really did end up internalising the lessons of the games, and that conclusion is borne out by the impact on their recidivism rates. What we did was far more honest and a damn sight more effective than anything that came before. Cheaper, too.”

Dr. Wood did harbour some doubts, though.

"Sometimes I wondered whether we were actually reforming offenders, or just teaching them to wear a mask that let them get by in society. I emailed my pastor with my worries, and he reassured me that only God could know for sure if a criminal was truly reformed. Since we can't look into someone's soul, he said that we could only look at how they behaved. He said that God would forgive us for any mistakes we made.

"I don't know whether he really believed that, or if he was only trying to make me feel better. I guess it doesn't matter either way, since I wasn't able to tell the difference."