Fascinating video - the distribution seems likely related to the HN post from two days ago on multiple random processes being sufficient to generate Zipf's law distributions [0] (given the observation a couple posts above that the first number not in the list is 18159 and the steady average decrease in frequency with value there isn't really much difference between rank-ordering and value ordering here). That exponent they find sure looks Zipf like.

Oh, and happy new year's!

[0] https://news.ycombinator.com/item?id=13281787