@dr_marcomeer@mastodon.green
Human dataviz is more likely to choose the right level of abstraction to make the most sense of the data. You do have to work about biases and distortions in the presentation, but the same is true of ML algorithms too, just more subtly so.