The "essentially static hosting" isn't the cost centre (although with 5 million MAU, it's nothing to sneeze at). The real costs are on the input side - they have an ingestion pipeline that ensures standardised paper formatting and so on, plus at least some degree of human review.
No, I mean that the pipeline requires software engineers to build/maintain, and salaries are (as in basically every tech organisation) the dominant cost
Make it an external service then, and leave the thing that's already working great to just be.
The reason authors like and use arxiv is that it gives 1) a timestamp, 2) a standardized citable ID, and 3) stable hosting of the pdf. And readers like the no-nonsense single click download of the pdf and a barebones consistent website look.
The PDF formatting is all but standardised. They ingest LaTeX sources, which is formatted according to the authors' whims (most likely, according to whatever journal or conference they just submitted the manuscript to).
I'll concede that the (relatively novel) HTML formatter gives paper a more uniform appearance. They also integrate a bunch of external services for e.g., citation metrics and cross-references. Still hard to justify such a high cost to operate, but eh.
Also, the "human review" is a simple moderation process [1]. It usually does not dig into the submission's scientific merits.
A critical component of the arXiv-CE project is moving our services entirely off of Cornell University’s infrastructure — this goal is also known as Milestone 1. Milestone 1 completion is projected for the end of fiscal year 2026.
Assume if you are a library, and every day, half baked so-called books brought to the librarians where they have to make sure it is meaningful, readable and printable, 3000 of them, they accept and put them in the right bookshelf, and entire internet reads every one of them on the shelf multiple times by the AI bots, search engines and researchers.
They are not only making a new library, they are also maintaining both and syncing two libraries because Cornell cannot handle the volume of access by bots.
It is not static. It is essentially running two ships side-by-side, and two ships need to appear as one from the outside. And, the new ship is still only half built. The new ship is being designed, and being built. 27 seems small to me.