> Not all computer vision is "low-hanging fruit," especially if you're pushing boundaries. Computer vision is low hanging in the sense that you already have ev

> Not all computer vision is "low-hanging fruit," especially if you're pushing boundaries. Computer vision is low hanging in the sense that you already have everything you need to make positive contributions to the study. I think the same is true with ML in general, but it was specific about what type of ML, which hangs higher due to the data availability. > Similarly, applying ML techniques to "other data that is relevant" is a far cry from using ML to save lives at a hospital due to misdiagnosis. Are you saying that programmers that are not working directly on saving lives are essentially wasting their time? There are a lot of interesting ML problems that do not save lives, but they are still worth working on.