Out of the 100 or so data scientists in my division, I know of less than 5 people who still occasionally use SAS. It's all either (decreasingly) R or (increasingly) Python.

edit: scala with spark is increasing in use as well.