Background

The framework scikit-learn uses predict for the application of model on (new) input data and I have seen many people use that term. In the scientific papers that I have read (and I read a lot, doing foundational research at university, but mostly in the area of unsupervised learning), the word predict seems not really that prevalent.

Looking up the word, it seems to me, that predict has a strong connotation about saying something about the future, but very often the prediction in the context of machine learning is not about the future at all. In fact, when people use a model to say something about the future, the word forecast is mostly used.

Additionally, the word predict seems to also have a connotation of soothsaying, which is also not what we are doing.

Because of that it seems a bit strange that the word predict has been used and thus my question:

Question

What I would like to know is the history of what lead the community / so many people / some frameworks to use the word predict. Maybe there is an original article that introduced the word, or an important talk or something like that. Maybe there was an influential discussion back in the day on why it is a good word that lead to the adoption.

I am also interested in arguments for the adoption of the word by influential people. So, I am not asking for an opinion of the one to answer the question (which again would be opinion-based), but references to and quotes of (historic) documents.