This is not a slight towards the OP, I’m just opening up for discussion. I think this is a great example of how we tend to default to algorithms and automation

This is not a slight towards the OP, I’m just opening up for discussion. I think this is a great example of how we tend to default to algorithms and automation even when the problem is better suited to a more manual solution. This is especially problematic when our automated systems don’t allow for transparent human intervention (think ML classifiers etc.). If this page were a curated wiki, for example, it’d probably have had a number of additions / corrections suggested by the community already. People are certainly posting corrections in this very thread, at least. Maybe the lesson is, use automated data as a seed and let humans iterate on that?