Lobster Roll

All Programming (General) 3773 Systems / Low-Level / OS 2950 Programming Languages / CS Theory 2127 Data / Databases / Infrastructure 1761 AI / Machine Learning 1522 Web Development 1308 Security / Privacy 1009 Culture / Philosophy / History / Reading 962 Productivity / Career / Business 731 Gaming / Retro Computing 538 Maker / DIY / Hardware 516 Design / UX / Visualization 491 Science / Math / Physics 428 Apple / macOS / iOS 355 Internet / Digital Culture 162 Health / Fitness / Lifestyle 14 Crypto / Blockchain 4

AI / Machine Learning

Stop Putting Secrets in .env Files (jonmagic.com)
Disrupting malicious uses of our [OpenAI] models (cdn.openai.com)
Instant LLM Updates with Doc-to-LoRA and Text-to-LoRA (pub.sakana.ai)
RetroTick, browser-based Windows emulator (retrotick.com)
FIPS: Free Internetworking Peering System (github.com)
AI sandbox that runs on your homelab (github.com)
Claude Code's 60fps Claim Is Perplexing (youtube.com)
Vibe-coding a debugger for a DSL (joachim-breitner.de)
Will vibe coding end like the maker movement? (read.technically.dev)
Poisoning AI Training Data (schneier.com)
Lessons from my overly-introspective, self-improving coding agent (ngrok.com)
Model Collapse Ends AI Hype (youtu.be)
Can LLMs SAT? (blog.aiono.dev)
On the Problem of LLM-assisted Contributions to Open Source Projects (call-with-current-continuation.org)
BuildKit: Docker's hidden gem that can build almost anything (tuananh.net)
Small Models Can Introspect, Too (vgel.me)
Mitchell Hashimoto’s new way of writing code (newsletter.pragmaticengineer.com)
Managing Complexity with Mycelium (yogthos.net)
The Agentic Mullet: code in the front, proofs in the back (amplifypartners.com)
Deterministic Programming with LLMs (mcherm.com)
Large-Scale Online Deanonymization with LLMs (substack.com)
Software Engineering Has Changed (lukesnotebook.substack.com)
How AI Will Change the Mobile Ecosystem (blog.bensontech.dev)
New accounts on HN 10x more likely to use EM-dashes (marginalia.nu)
Current Large Audio Language Models largely transcribe rather than listen (arxiv.org)
Tests Are The New Moat (saewitz.com)
Mercury: Ultra-Fast Language Models Based on Diffusion (arxiv.org)
I Taught My Dog to Vibe Code Games (calebleak.com)
ansigpt: c89 implementation of microgpt (github.com)
RFC 406i - The Rejection of Artificially Generated Slop (RAGS) (406.fail)