I discourage use of machine learning, because it’s really difficult. Unless you have to use machine learning, don’t do it. It’s usually a red flag when somebody is saying “We wanna use machine learning to solve this task”. I’m like: That sounds like bullshit. 99.9% of the time you don’t need it.
– Elon Musk
That’s a statement made by Elon Musk during past weeks “Tesla AI Day” when being asked if Tesla also uses machine learning for other applications than autonomous driving.
It sums up pretty nicely what I think about useful use cases for machine learning as well. I believe there is often a much more simpler and faster way to achieve a desired result. Machine learning done right is hard and requires a lot of knowledge, training data and compute capacity to get right. Otherwise you might end up with an unfit model whose output looks good, but actually isn’t, which might not be obvious as the reasons for decisions made by the model are too opaque.