frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

Open in hackernews

Implementing WebNN with the Help of AI

https://medium.com/@polyglot_factotum/sketching-webnn-with-ai-part-two-of-the-slop-diaries-8df62d4102ca
1•polyglotfacto•1h ago

Comments

polyglotfacto•1h ago
I am a maintainer of the Servo web engine. Two years ago, our project banned AI-generated contributions. I was a dissenting voice on that policy, and have continued to experiment with AI on side-projects. This is perhaps the most ambitious side-project so far: implementing the WebNN (Neural Network) API.

Given the policy in place, I'm doing this on a personal branch of Servo. I’m using it as a sandbox to test a specific thesis: AI is not an architect; it is a "fluid syntax engine."

This is also my response to the "autonomous slop" seen in projects like fastrender that claim to build complex software with AI but produce what I call "fine-looking nonsense" (see https://news.ycombinator.com/item?id=46624541#46709191). I wanted to see if a human architect could use an AI agent to handle the syntax and boilerplate of a Web Standard while the human retains total control over the conceptual structure.

Progress & Tech Stack:

    The scope: I am using `rustnn`, and this project can be seen as doing for the Web what `pywebnn` does for Python. 

    The model: I am intentionally using a "cheap" LLM: Raptor mini.

    The scale: I’m 7k LOC into the implementation. It currently passes a subset of WPT (Web Platform Tests) conformance tests for a specific operator (and its dependencies): add. Such a conformance test requires implementing the full graph compilation and dispatch workflow. Code is at https://github.com/gterzian/servo/compare/master...gterzian:servo:webnn

    The backend: The work involves integrating the `rustnn` library into Servo’s runtime, with compute currently happening only via CoreML (macOS only).
The Reality Check: While the productivity boost is real, the code is not yet up to my standards. Despite ongoing reviews, I’ve had to accept a certain amount of "slop" that I’ll need to clean up later. More importantly, I hit several conceptual bottlenecks: architectural problems the AI was not only unable to solve without guidance but was unable to even identify in the first place.

You can read the full breakdown of the first week and a half (contains links to code and various illustrative commits) in the link for this post.

And the backstory on why I'm doing this is at: https://medium.com/@polyglot_factotum/the-slop-diaries-imple...

I’d be happy to hear from others on how they balance the "syntax speed" of AI with the "architectural integrity" required for long-term projects.

A playbook for the next century of progress

https://www.freethink.com/opinion/techno-humanist-manifesto-chapter-11
2•bryanrasmussen•6m ago•0 comments

Show HN: Bunny Inspector – Show your AI what to fix – in one click

https://inspector.bunnytech.app
1•iosifnicolae2•7m ago•0 comments

Peer Pressure Works on AI Too

https://www.robkopel.me/field-notes/peer-pressure/
1•umanghere•9m ago•1 comments

AOMedia Open Audio Codec "OAC" Aims to Be the Successor to Opus

https://www.phoronix.com/news/AOMedia-OAC-Open-Audio-Codec
3•breve•9m ago•0 comments

Claude Code Remote Control

https://twitter.com/claudeai/status/2026418433911603668
2•mdrachuk•10m ago•2 comments

INSIDE 2016

https://www.entertainmentbreakdown.com/playdeads-inside-10-years-anniversary-review-e31381c806bf
1•raynchad•11m ago•0 comments

The Manifold of Breakfast Has Gaps

https://moultano.wordpress.com/2026/02/22/the-hunt-for-dark-breakfast/
1•moultano•11m ago•0 comments

Hacking Claude Code remote: escaping YOLO-mode sandboxing

https://www.noahlebovic.com/hacking-claude-code-on-the-web-breaking-yolo-mode-sandboxing/
2•lebovic•12m ago•0 comments

Claude Cowork starts rolling out scheduled tasks

https://old.reddit.com/r/ClaudeAI/comments/1re1ncv/did_anyone_else_notice_cowork_now_has_scheduled/
1•mdrachuk•14m ago•1 comments

Tell HN: Planning to end my life today

2•throwawayxdsds•15m ago•1 comments

Obsessed with Terminology

https://faidi.tech/posts/terminology/
1•ghassenfaidi•18m ago•0 comments

Show HN: Roundsman – stupid-simple CLI to run Claude across many projects

https://github.com/tsortwehttam/roundsman
1•matthewtoast•18m ago•0 comments

AI Rage: Why it happens and what to do about it

https://github.com/chrhartm/ai-rage/blob/main/README.md
2•christoph123•19m ago•0 comments

OpenClaw deletes Summer Yue's emails

https://twitter.com/summeryue0/status/2025774069124399363
2•teekert•20m ago•0 comments

Last Year of Terraform

https://encore.dev/blog/last-year-of-terraform
6•eandre•23m ago•2 comments

The Emancipated Codebase: Why AI Just Fired Its 1970s Babysitter

https://the-mind-of-ai.com/posts/emancipated-teenager/
1•agentic-wiki•24m ago•1 comments

Nano Banana 2 is real!Gemini 3.1 Flash Image just appeared in Vertex AI Catalog

1•zaaaaooo•24m ago•0 comments

"Affordances," a science-fiction story by Cory Doctorow about algorithmic bias

https://slate.com/technology/2019/10/affordances-cory-doctorow-sf-story-algorithmic-bias-facial-r...
1•rbanffy•26m ago•0 comments

Show HN: SnapSift – Find near-duplicate photos on your iPhone

https://snapsift.app/
1•anhphong•26m ago•1 comments

AI Coding Agents Modify Code

https://arxiv.org/abs/2601.17581
2•nkko•28m ago•0 comments

Show HN: Hour Cap – From timer to Xero invoice in one click

https://hourcap.com
2•jonobird1•28m ago•0 comments

What breaks when AI agent frameworks are forced into <1MB RAM and sub-ms startup

3•NULLCLAW•28m ago•2 comments

Best MCP Servers for Knowledge Bases

https://desktopcommander.app/blog/best-mcp-servers-for-knowledge-bases-in-2026/
1•rafaepta•30m ago•0 comments

Goblin Supermarket – Closing Time

https://medium.com/luminasticity/goblin-supermarket-part-iii-fb58504d3dd7
1•bryanrasmussen•31m ago•0 comments

Designed for Two: The Co-Op Philosophy of Josef Fares

https://www.brothers-in-gaming.com/post/designed-for-two
2•vinnyita•32m ago•1 comments

Show HN: A human-curated, CLI-driven Context Layer for AI agents

https://ctxlayer.dev/
2•anatoliikmt•33m ago•0 comments

Part II. SCM as a database for the code

https://replicated.wiki/blog/partII.html
2•gritzko•35m ago•0 comments

Show HN: MCP-enabled file storage for AI agents, auth via Ethereum wallet

https://service.c33b.org/
2•womd•35m ago•0 comments

The "Electron-ification" of Windows is killing my Surface Book 2

https://old.reddit.com/r/Windows11/comments/1r5bkl1/the_electronification_of_windows_is_killing_my/
6•thunderstruck•38m ago•1 comments

Building Aira: a calm anime tracker for myself

https://blog.kulman.sk/building-aira/
2•ig0r0•38m ago•0 comments