I really enjoyed the writing style in this article.

And the bot progression from "alter user agent" to "change IP addresses" to providers having to ban whole subnets, whole ASNs, and realizing "proxy SDK monetization" is a thing mirrors threat actor progression from the time before LLMs.