







No training anywhere: a compiler sets the weights of a stock Phi-3-architecture checkpoint so autoregressive generation executes Doom's classic renderer.
Noel Welsh: Doom Emacs Workflows
I recently switched to Doom Emacs from Spacemacs. The Doom documentation is currently quite sparse, so I've posted here my personal notes on using Doom along with a description of my workflow (something I find vital but missing from most documentation). Doom might be for you if
The Virtual Reality of John Carmack
The Dallas video game genius behind “Doom” and “Quake” is back, this time in the nascent VR space. In this exclusive interview, he tells why.

Production Is a Compiler Input - The Phoenix Architecture
The Generative Stack - The Phoenix Architecture
Trying to find the best tool or platform for generative software in 2026 is a mistake that could haunt you for decades
EXCLUSIVE---DOOM: The Dark Ages Revelations Faced ‘Worst Crunch’ in Franchise History
DOOM: The Dark Ages Revelations faced brutal crunch, with developers working up to 80-hour weeks to finish the expansion before release.

Self-Consuming Generative Models Go MAD
Seismic advances in generative AI algorithms for imagery, text, and other data types has led to the temptation to use synthetic data to train next-generation models. Repeating this process creates an autophagous (self-consuming) loop whose properties are poorly understood. We conduct a thorough analytical and empirical analysis using state-of-the-art generative image models of three families of autophagous loops that differ in how fixed or fresh real training data is available through the generations of training and in whether the samples from previous generation models have been biased to trade off data quality versus diversity. Our primary conclusion across all scenarios is that without enough fresh real data in each generation of an autophagous loop, future generative models are doomed to have their quality (precision) or diversity (recall) progressively decrease. We term this condition Model Autophagy Disorder (MAD), making analogy to mad cow disease.

Doomscrolling
Doomscrolling is the act of spending an excessive amount of time on digital content that elicit negative emotions. The concept was coined around 2018, and became more widespread in the context of the COVID-19 pandemic. The World Health Organization (WHO) observed that the pandemic was accompanied by widespread misleading information, conspiracy theories, and false reports, which it referred to as an "infodemic".
Getting started with Doom Emacs — A great transition from Vim to Emacs
So you just installed Emacs(make sure it’s 26+!) and you want to see what all the fuss is about. Why should I use this instead of my…

The AI Doomers Have Infiltrated Washington
This week, a State Department funded report declared that AI posed an existential threat akin to nuclear weapons—and even recommended jail time for unauthorized AI development.

Compiling Models to Megakernels
Fine-grained synchronization, deep pipelines, and zero kernel launch overheads, automatically.

Warp — Open Infrastructure for Cloud Software Factories
Build your software factory on Warp: factories as code, any model or harness, with evals, benchmarks, and self-improvement built in.

Fake Ass Spec
This is a fake-ass spec. We can annotate things1. We can link. We don't believe in P(doom^k). But we can keep on going!

Visualizing GL_NV_shader_sm_builtins
Using GL_NV_shader_sm_builtins to visualize Streaming Multiprocessors and Warps
AI Doomsday Bullshit Is Getting Tired
The "rogue AI" narrative continues to help obscure very ugly and avoidable human failures.
I made a Doom source port that runs within a PDF file. PDFs support Javascript, so Emscripten is used to compile Doom to asm.js, which is then run within the PDF engine. Input/output is done by manipulating text input fields. doompdf.pages.dev/doom.pdf github.com/ading2210/doompdf