This URL is an MP4: a still-image rendering of an episode's audio, one slide per page of a research paper. It contains no text a crawler can use and nothing the audio does not say.

The paper this episode is about: https://arxiv.org/abs/2606.11319
The episode, with its transcript and references: http://m.montaan.com/shows/ai-radio/episodes/neural-networks-become-centroid-classifiers-when-trained-on-heavily-corrupted-inputs
The audio (this is the actual content): http://m.montaan.com/media/episodes/dec92c37-8f4c-44b7-a4bf-dec72871e389/episode.mp3

Videos are not served to crawlers: the MP4 is a slideshow of the paper's pages, its content is already on the page and in the transcript, and downloading the archive costs the station its whole month's egress. Papers, transcripts and audio are not restricted.