I think Kristine has done a really great job narrating and reflecting on her experience using the Detain/Release simulator. She has included a play by play of her actions and also included some reflection questions, right at the start, to get the reader thinking. As Kirstine notes, using an algorithm in isolation is challenging, especially when dealing with such sensitive information in a highly-consequential situation. She elaborates that the trade off of efficiency is not always the right approach, especially given the narrow data-driven approach of algorithms. 

One of the key points Kristine outlines is, “For each final decision I made, I wondered if it might result in a bigger consequence such as homicide, rape, loss of a home or job, or even death by suicide.” One of the first things that came to my mind was the possibility of sending a violent man back into the world. Why? Because I too was also concerned about rape, sexual violence, and domestic abuse. Clearly, indicating that I had a bias towards men as I sorted through the cases. As a woman, I move through the world, on some level, afraid of men. Covering my drink, taking out my headphones, looking over my shoulder, walking away from the road, etc… are all things that change/impact my life because I am afraid a dangerous man will put my well-being at risk. Throughout this experiment, that sentiment lingered in the back of my mind. Subconsciously thinking, how likely is it that a woman is going to go home and abuse her spouse? But with so limited information, making an honest and accurate assessment is challenging – even with the right resources we can’t always know what’s going on behind closed doors. This simulation certainly forced me, and likely many of its players, to think about their underlying biases.