Why do random forests work? They are self-regularizing adaptive smoothers

Reading the abstract alone, I have no idea whether it's talking about algorithmic trees or, like, the big brown things with small green bits.

The tree is an incredibly common data structure in computer science. Decision trees are well known. Random forests are ubiquitous in Machine Learning. Should the authors really have to dumb their paper down so people who don’t work in this domain avoid confusing it with work in arborism?