You ever notice how a perfectly optimized algorithm can be both a blessing and a curse? I’ve been looking into that with the new predictive policing model. What do you think about tightening the parameters?