Essay called The Education of a Doomer turns out to be about AI, not id's Doom

A widely shared essay with a Doom-adjacent title has nothing to do with id Software's engine. It is an account of one engineer souring on AI.

AI-drafted from 2 sources. This story was written with AI from the outside reporting listed under Sources, and checked by automated tools before it went up. The originals have the full detail.

An essay titled “The Education of a Doomer” has been doing the rounds on Hacker News this week, filed under a thread tagged emulation. Anyone hoping for a technical retrospective on id Software's Doom engine, the one that gave the world BSP trees and speedrunning culture, will be disappointed. The piece, written by Fernando Borretti, is not about the 1993 shooter at all. It is a personal essay about why its author went from being “generally optimistic and excited about AI” to “extremely concerned” about it, and the “Doomer” in the title is the internet slang for someone who believes AI poses existential risk, not a nod to Doomguy.

It is an easy mistake to make. The Hacker News discussion thread sits under an emulation tag, and the title reads, at a glance, like it belongs on a blog about classic engine architecture. It does not. Borretti structures the essay as a series of sections, among them economics, alignment, control, writing, software engineering, deferring to AI and stagnation, each one describing a belief he used to hold about artificial intelligence and why he no longer holds it.

What the essay is actually about

The argument that will likely interest readers of a gaming and software site is the one filed under “Writing” and “Software Engineering.” Borretti writes about opening pull requests on his own projects and finding “the code is slop,” about README files and blog posts that read the same way, and about a college professor defending the use of AI to write academic papers under their own name. He connects this to a broader worry that reinforcement learning, which he says has been “the main technique to push the frontier forward” since roughly 2024, produces models that are more capable but harder to understand, because, in his words, “their prose is increasingly incomprehensible.”

He frames this against an older, more optimistic view he once held: that automation has historically made people richer rather than idle, citing the wholesale disappearance of 1790s-era jobs without mass unemployment as a reason to expect something similar this time. The essay's newer position, laid out across sections with titles like “No-One Escapes the Permanent Underclass” and “Our Servants Will Do That For Us,” is considerably darker: that if AI becomes both general and dramatically cheaper than human labour, there may be nothing left for people to trade for a stake in political power, UBI included.

Why this ended up near a games audience at all

The honest answer is a title pun and a forum tag, nothing more. There is no BSP tree in sight, no discussion of Doom's renderer, no engineering lesson about the game that shaped a generation of shareware. The essay is a first-person account of one programmer's change of mind about artificial intelligence, written for a general audience rather than a games one, and it happens to have surfaced on a forum thread whose tagging suggested otherwise.

What is worth taking from it, if anything, for developers reading Hacker News or GBAtemp, is the specific complaint about pull requests and documentation quality, since that is the part grounded in something concrete rather than speculative. Borretti does not offer numbers or studies for that claim, only his own experience reviewing code and papers, so it remains anecdotal. Readers looking for the Doom engineering essay this headline seems to promise will need to look elsewhere, because that essay, as far as the source itself shows, does not exist here.

Sources

• The Education of a Doomer (Fernando Borretti)
• Discussion: The Education of a Doomer (Hacker News)

Image: Fernando Borretti