Actually it was an unused small Vax (730?) running BSD. It was unused because it was in the AI lab, where we all used lispms or the PDP-10s upstairs (don't get me wrong -- some of the non-AI folks in computer science used unix).
I don't even remember why that machine was there in the first place -- it seemed so peculiar to have a machine like that in the first place. 8-bit bytes and the small, 32-bit word didn't feel like the future at the time (hah!). It was stuck in a small machine room.
(I wasn't running GNU (emacs) until 1984, and didn't contribute to GCC until 1987.)
Good thing for my self image I put that question mark in there.
DEC had a problem with selling MIT and other non-profits hardware at a discount, the US Federal government demands Most Favored Customer prices (https://en.wikipedia.org/wiki/Most-Favoured-Customer_Clause), so I heard they came up with a not so desirable configuration of a Vax 11-750, a RK07 disk drive, and a VAXstation 100 display, which was a 68000 with a megabyte or so of memory and a very fast connection to the VAX.
Which I was also told had an outsized influence on the X architecture, its assumptions of a fast connection to a relatively dumb display, very much the opposite of what developed in longer term.
Which was a motivation for Sun's NeWS, which sent interactive PostScript to the display https://en.wikipedia.org/wiki/NeWS We aren't using it/you've never heard of it because of the "MIT License" vs. Sun wanting $100K to even give you a taste of NeWS.
It is not that his doing didn't turn out to be intelligent, but his intent almost certainly wasn't.
This isn't incompatible with earnestly believing the views. All of us hold principles that exist on a spectrum and can pragmatically decide how we apply them without necessarily compromising them, in order to optimize for progress of certain kinds on certain time-frames with certain costs.
An example of this, RMS's positions have become stronger as free software as matured: IIRC GNU was originally developed on Sun boxes; but RMS would never use an outright proprietary OS today. Early free software system he used had non-free firmware or other blobs, but he advocates strongly against that today. There are a set of strong ideals, sure, but the amount of compromise someone holding them should optimally take depends on the overall state of the world and their position to influence it.
The notion of software freedom as a human rights issue can be quite foreign to people and without someone taking it very seriously some, especially those who have mostly regarded technology as a black box, can too easily dismiss it.