I think that the most surprising piece of information in this discussion was the comment about trusting predictions. When we talk about predictions, sometimes we have to choose different models of what we believe or trust, or sources of information that we tend to rely on more. But, it could, in fact, be more about trusting a prediction that is not 100% correct all of the time, so long as it is more accurate than other predictions. This principle, in and of itself, relates to my final project because it is important to consider the uncertainty around election predictions. The models of different websites, like AP News or NYT or WSJ and many more are all different. But as a voter, it is your choice to make a decisions about what model you trust. Obviously, regardless of the prediction model that you choose to trust, there is always going to be one outcome of the election, whether or not you are incorrect in the model that you chose (and it was wrong about the outcome). Still, in order for you to buy into a predictive system, you just need to choose the one that you think is best, since it is still better than baseline.
Here is a NYT map of the electoral outcomes in the United States. This represents one of the models and data types (historical) that you could use to make a prediction about the outcome of a future election. https://www.nytimes.com/interactive/2021/upshot/2020-election-map.html
Interview Discussion: