Detain/Release Linking
In response to Leilani’s detailed and thoughtful insights on the Detain/Release task I commented on her post as follows:
Hi Leilani,
Reading your post gave me perspective as to why I was unable to complete the task successfully with the methods and criteria that I chose. What was likely a key part of my failing (and your success) was that I did not recognize as accurately or as quickly as you did that “those who [you] thought had committed crimes that were not as “bad” were the ones that fled”. We had a similar stance on weighing the severity of the crime in relation to violence but I suspect you may have adjusted your actions as people began to flee more quickly than I did. You describe that you became trigger happy and I did as well but described it more as a feeling of being up against a clock. While we seem to have similar approaches we ended with different results. It might be that you identified that the AI targeted those in poverty to a greater degree which as you state has links to marginalized communities and adjusted your methods of ruling on a crime. The one thing that I still struggle with this task is that jail capacity and fear are both quite high (I would say unacceptably so) but this is the reality unless we consider addressing the source of the crime (poverty) as you point out.
Thank you,
Chris
Reason for linking:
Until I read Leilani’s post I was not quite convinced it was possible to succeed at this task but her post proved it was not only possible, but it was possible to do so without the fear bar entering into the red zone. While we had similar approaches to this task she made a few key observations more quickly than I did in relation to those who fled, which I suspect led to her success. After completing the task I wanted to run the simulation again to test different hypotheses I came up with after completing my first task. I resisted because we were not invited to do so specifically. As such I appreciate that Leilani’s post provided insight into the questions I had after completing the simulation.