Anyone else get brain lag from overly-condensed AI-speak? [D]

We’ve all had the experience of coming across a phrase like “quiet single-source substitute” which just barely makes sense in the context of the conversation but takes way too much effort to parse. What research exists into making AI more readable?