HMML: A Conceptual Framework for Composable Document Generation Over Static Pixels
By
yeargun
Summary
This article introduces HMML (HyperMedia Markup Language), a conceptual framework that proposes generating composable documents instead of static pixel-based images. The core idea is that AI models should output structured, editable elements (vector, text, raster, 3D, motion) at the node level rather than flattening everything into a frozen raster grid. The article argues that "image" is an overloaded term and HMML offers a way to un-conflate it, though it remains speculative about whether models will adopt this approach.
Source
Hacker NewsHMML: A Conceptual Framework for Composable Document Generation Over Static Pixelshmml.eddocu.comKey quotes
· 5 pulledcomposability, not pixels
The next thing a model generates isn't an image. It's a document.
An image flattens everything into one frozen raster. HMML keeps the pieces - vector, text, raster, 3D, motion - composable and editable, created at the grain of a node, not a 1024-grid of guesses.
But 'image' is already an overloaded word - photo, icon, chart, scene, animation, all crushed into one frozen raster.
Or maybe pixels were fine all along. Let's see.
You might also wanna read
Why HTML is the Ideal Language for AI Agents to Create Visual Content
Amol Kapoor, CEO of Nori Agentic, argues that HTML is the ideal universal language for AI agents to create visual content like slides, docum
startuphub.ai·3h ago
LoomVideo: A 5B-Parameter Unified Model for Efficient Video Generation and Editing
LoomVideo is a new 5-billion parameter unified architecture for video generation and editing that addresses computational bottlenecks in exi
arxiv.org·22d ago
JAMEL: A Framework for Joint Memory and Exploration Learning in Language Model Agents
This paper introduces JAMEL (Joint Agent Memory and Exploration Learning), a framework that trains language model agents to explore open-end
arxiv.org·23d ago
ShapeLib: Using LLMs to Design Programmatic 3D Shape Abstraction Libraries
ShapeLib is a novel method that leverages Large Language Models (LLMs) to design libraries of programmatic 3D shape abstractions. The system
arxiv.org·26d ago
LOGOS: A Unified Generative Foundation Model for the Natural Sciences Using Shared Scientific Grammar
This report introduces LOGOS (Language Of Generative Objects in Science), a scientific generative language model that unifies diverse tasks
arxiv.org·11d ago
AI-First Content Management: Rethinking CMS vs Markdown for Agentic Applications
The article explores whether traditional Content Management Systems (CMS) like WordPress are still necessary in an AI-first world where agen
Prototypr·5mo ago
Comments
Sign in to join the conversation.
No comments yet. Be the first.