Speaking of their API, a recently new one is the Archive API (I only noticed it this past year), which makes it much easier to download all of their article metadata. It used to be you had to paginate through their articles API, which was limited in its number of paginated results. Now you can get all of the articles and their metadata for any given year and month, which makes it a great place for practicing data mining and other classification techniques.

https://developer.nytimes.com/archive_api.json#/README

Example result for 2016-11 (around 18mb in size):

http://stash.compciv.org/2017/nytimes-archive-2016-11.json